Combined therapy with recombinant granulocyte colony-stimulating factor and erythropoietin decreases hematologic toxicity from zidovudine.

Twenty-two patients with acquired immunodeficiency syndrome (AIDS) or severe AIDS-related complex and multilineage hematopoietic defects were treated with recombinant granulocyte colony-stimulating factor (G-CSF) and erythropoietin (EPO) in a phase I/II trial. All patients were neutropenic and anemic after withdrawal of all bone marrow-suppressive drugs. Daily, G-CSF was subcutaneously self-administered until an absolute neutrophil count (ANC) greater than 6,000/microL was achieved and maintained for 2 weeks. Subcutaneous EPO was added to the regimen and the dose increased until an increase of 15 g/L of hemoglobin was observed. Groups of patients were administered increasing doses of zidovudine to determine their tolerance. G-CSF and EPO therapy was continued with dose modification to maintain an ANC greater than 1,500/microL and hemoglobin greater than 100 g/L. The dose of zidovudine was not altered. All 22 patients responded to G-CSF with a mean 10-fold increase in neutrophils occurring in less than 2 weeks. Significant increases in CD4 and CD8 cell number, lymphocyte proliferative response, and bone marrow cellularity were seen. EPO therapy increased hemoglobin in all 20 evaluable patients within 8 weeks. Sixteen patients received 1,000 mg and four patients received 1,500 mg of zidovudine per day. The reinstitution of zidovudine resulted in a decline in reticulocytes and hemoglobin and the reappearance of transfusion requirements in eight of the 20 patients, six of whom had the study medications stopped. No patient had the study medications stopped because of neutropenia or thrombocytopenia. Toxicities were mild and did not require dose modifications. Limiting dilution plasma and lymphocyte co-cultures for HIV as well as serum p24 antigen levels did not change significantly during G-CSF or combined G-CSF and EPO therapy. HIV p24 antigen decreased significantly with zidovudine therapy. Opportunistic infections occurred in 14 patients but were successfully treated with myelosuppressive antimicrobial agents, including ganciclovir, without the development of neutropenia. These results suggest that combined therapy with G-CSF and EPO may improve the neutropenia and anemia of AIDS. Combined therapy may allow the resumption of full-dose zidovudine in most patients intolerant of the hematologic effects of zidovudine without apparent alteration of HIV expression or the efficacy of zidovudine.

Clinical studies with granulocyte colony stimulating factor (G-CSF) in patients receiving cytotoxic chemotherapy.

Bacterially synthesised human granulocyte colony-stimulating factor (G-CSF) was administered to patients with advanced cancer. The immediate effect of G-CSF was a fall in the level of circulating neutrophils followed by a rise after 4 hours that was sustained during G-CSF administration. The rise in neutrophil level was less in patients who had been treated previously with chemotherapy and/or radiotherapy. G-CSF was also administered to patients following melphalan and this resulted in a reduction in the duration of the neutropenia that invariably follows melphalan. G-CSF was well tolerated and did not have to be stopped in any patient.

Phase I/II study of recombinant human granulocyte colony-stimulating factor in patients receiving intensive chemotherapy for small cell lung cancer.

Twelve patients with advanced small cell carcinoma of the bronchus were treated by continuous infusion of recombinant human granulocyte colony-stimulating factor (rh G-CSF) at the following dose levels: 1 microgram, 5 micrograms, 10 micrograms, 20 micrograms and 40 micrograms/kg/day for 5 days. No toxicities resulted from the treatment and in all 12 patients the number of peripheral neutrophils increased rapidly to a maximum of 100 x 10(9)/l in one patient at 10 micrograms/kg/day. The neutrophils were shown to be functionally normal in tests of their mobility and bactericidal activity. During the Phase II part of the patients were treated using a combination of i.v. Adriamycin, Ifosfamide and Etoposide. The chemotherapy was repeated every 3 weeks. rh G-CSF was given to each patient for 14 days on alternate cycles of chemotherapy and reduced the period of absolute neutropenia considerably (median of 80%), with a return to normal, or above normal, neutrophil counts within 2 weeks after day 1 of chemotherapy. Ten severe infective episodes were observed during the 20 cycles of chemotherapy which did not include rh G-CSF, while only one infective episode occurred in 20 courses when treated with rh G-CSF. These results demonstrate the utility of rh G-CSF in restoring functional neutrophils to patients undergoing intensive chemotherapy.

In vitro and in vivo analysis of the effects of recombinant human granulocyte colony-stimulating factor in patients.

Twelve patients with small cell lung cancer were treated with recombinant human granulocyte colony-stimulating factor, rhG-CSF, given by continuous infusion at doses ranging from 1 to 40 micrograms kg-1 day-1. Patients received the rhG-CSF before the start of intensive chemotherapy and after alternate cycles of chemotherapy. Several in vitro assays were performed using peripheral blood neutrophils and marrow progenitor cells collected from patients prior to and after infusion of the growth factor. Peripheral blood neutrophils were tested for mobility and phagocytic activity. In addition, in vitro clonogenic assays of marrow haemopoietic progenitor cells and analysis of bone marrow trephines and aspirates were carried out. We found that rhG-CSF in vivo has at least two main effects: (a) an early fall in peripheral neutrophils, within the first hour, followed by a rapid influx of mature neutrophils into the circulatory pool; (b) stimulation of proliferation and differentiation of neutrophil precursors in the bone marrow. Neutrophils released into the circulation were normal in tests of their mobility and phagocytic activity.

Phase I study of recombinant methionyl human consensus interferon (r-metHuIFN-Con1).

Consensus interferon (r-metHuIFN-Con1) is the product of a gene constructed to code for the most frequent amino acid residues known to occur in subspecies of alpha interferons. Twenty-one patients with advanced malignancy entered this phase I trial with dosing levels of 3, 7.5, 15, 30, and 45 mcg/m2/day given intramuscularly on days 1-5 and 8-10 of each 28-day cycle. The initial dose was randomly given by intravenous, intramuscular, or subcutaneous injection to facilitate pharmacokinetic studies. Vomiting and diarrhea were dose-limiting at 45 mcg/m2/day, preventing completion of therapy. Malaise, flu-like symptoms, nausea, and headache were frequent but tolerable at a dose of 30 mcg/m2/day. Patients were able to escalate to 45 mg/m2/day, suggesting tachyphlaxis to these toxicities. The initial distribution phase (T1/2 alpha) was 4.9-9.0 minutes with a T1/2 beta of 34-415 minutes in three patients for whom sequential values could be determined. r-MetHuIFN-Con1 was absorbed after both subcutaneous and intramuscular administration. 2'5'-Synthetase levels increased following treatment, although no consistent pattern was noted. One partial response was seen in a patient with gastrointestinal carcinoma. The recommended phase II starting dose of r-metHuIFN-Con1 is 30 mg/m2/day using this schedule by any of these routes of administration.

A phase I trial of recombinant human gamma interferon (IFN-gamma 4A) in patients with advanced malignancy.

We report a Phase I study in 39 cancer patients of the tolerance and biologic activity of 47 intravenous (i.v.), intramuscular (i.m.), and subcutaneous (s.c.) treatments with recombinant methional gamma interferon (IFN-gamma 4A) which most closely resembles the natural material produced by T lymphocytes. Patients were treated with IFN-gamma 4A 5 days a week for 2 weeks. After a 2-week rest period, patients were placed on the same dose of drug three times a week. The most common side effects--fever, chills, malaise, myalgias, and nausea and vomiting--were seen with all routes of administration. Reversible increases in hepatic transaminase and decrease in granulocytes counts were seen. The dose-limiting toxicities observed were malaise and orthostatic hypotension. The maximum tolerated dose was 500-1,000 micrograms/M2/day. The t1/2 of IFN-gamma 4A in the circulation was 20 min after i.v. injection. No blood levels were detected after i.m. or s.c. injection. Antibody against IFN-gamma 4A increased in three patients. A complete response was observed in one patient with pulmonary metastases from renal cell carcinoma.

Effect of granulocyte colony stimulating factor on neutropenia induced by cytotoxic chemotherapy.

A phase I/II study of granulocyte colony stimulating factor (G-CSF) was undertaken in patients with advanced malignancy receiving melphalan to determine the granulocyte response, side-effects, and pharmacokinetics. Patients received doses of 1-60 micrograms/kg intravenously. There were 3 patients at each dose level. Before chemotherapy the immediate effect of G-CSF was a transient depression in circulating neutrophils followed by a dose-dependent rise. Neutrophil counts up to 80 X 10(9)/l were achieved. G-CSF administration following melphalan reduced the period of neutropenia caused by melphalan. G-CSF was well tolerated and the only clinical observation that appeared related to G-CSF administration was slight bone pain during some infusions. G-CSF was rapidly cleared from the blood with a mean half-life of 110 min for the second phase. Reductions in the number of days of neutropenia following cytotoxic chemotherapy may reduce the morbidity and mortality of chemotherapy.

Phase I/II study of recombinant human granulocyte colony-stimulating factor in patients receiving intensive chemotherapy for small cell lung cancer.

Twelve patients with advanced small cell carcinoma of the bronchus were treated by continuous infusion of recombinant human granulocyte colony-stimulating factor (rhG-CSF) at the following dose levels: 1 microgram, 5 micrograms, 10 micrograms, 20 micrograms and 40 micrograms kg-1 day-1 for 5 days. No toxicities resulted from the treatment and in all 12 patients the number of peripheral neutrophils increased rapidly to a maximum of 100 x 10(9) l-1 at 10 micrograms kg-1 day-1. The neutrophils were shown to be functionally normal in tests of their mobility and bactericidal activity. During the phase II part of the study the patients were treated by a combination of intravenous adriamycin 50 mg m-2, ifosfamide 5 g m-2 by i.v. infusion with mesna 8 g m-2 on day 1, and etoposide 120 mg m-2 on days 1, 2 and 3 also intravenously. The chemotherapy regime was repeated every 3 weeks. RhG-CSF was given to each patient for 14 days on alternate cycles of chemotherapy and reduced the period of absolute neutropenia considerably (median of 80%), with a return to normal, or above normal, neutrophil counts within 2 weeks after day 1 of chemotherapy. Six severe infective episodes were observed during the cycles of chemotherapy which did not include rhG-CSF, while no infective episodes occurred when patients were treated with rhG-CSF. These results demonstrate the utility of rhG-CSF in restoring functional neutrophils to patients undergoing intensive chemotherapy.

Conformation and stability of two recombinant human interferon-alpha analogs.

Structural features of a recombinant E. coli derived interferon-alpha analog, interferon consensus1, was studied by circular dichroism and fluorescence spectroscopy. Circular dichroic spectra of the purified protein showed that it has about 70% alpha-helix and a distinct tertiary structure. These structural features are similar to those for a natural interferon-alpha subtype, interferon-alpha 2, indicating that the amino acid substitutions in interferon consensus1 apparently did not alter the protein structure. Another analog, interferon consensus5, which has Ser instead of Cys at residues 1 and 99 but is otherwise identical to interferon consensus1, was prepared to study the role of the disulfide bond between Cys 1 and 99. Circular dichroic and fluorescence spectra indicated similarity in the structure of these two analogs. However, interferon consensus1 was significantly more stable than interferon consensus5 against denaturation. pH unfolding experiments indicated that the former protein is more stable in the transition region by about 1.6 kcal/mol, which was interpreted in terms of the increased free energy of the denatured state due to an extra disulfide bond in interferon consensus1.

Structure of human tumor necrosis factor alpha derived from recombinant DNA.

Recombinant DNA derived tumor necrosis factor alpha, when expressed at a high level in Escherichia coli, appeared in the pellet and soluble fractions of disrupted cells. The protein was purified from the pellet fraction by solubilizing it in urea and reducing agent and was refolded into a buffer without these additives. The structure of the protein was identical with that purified from the soluble fraction without exposure to both reducing and denaturing agents, as demonstrated by circular dichroism, gel filtration, and sulfhydryl titration. As a reflection of the structural similarity, both purified proteins showed identical cytolytic activity on mouse L929 cells. The protein was characterized as an essentially nonhelical and beta-sheet-rich structure and possibly as a noncovalently associating oligomer. Two cysteine residues form an intrapolypeptide disulfide bond.

Structure and activity of recombinant human interferon-gamma analogs.

We have prepared interferon-gamma (IFN-gamma) analogs to study the structural role of particular amino acids in relation to their effects on antiviral activity. Three IFN-gamma analogs were prepared on the basis of predicted secondary structure. In two of the analogs, [Gln25]IFN-gamma and [Thr45]IFN-gamma, changes were made at residue 25 (Asn to Gln) and at residue 45 (Met to Thr), respectively. [Gln25Lys78]IFN-gamma had two changes, at residue 25 (Asn to Gln) and residue 78 (Asn to Lys). Another analog, [Cys-Tyr-Cys]IFN-gamma, incorporated Cys-Tyr-Cys at the amino terminus. Comparison of the structure and activity of these analogs with that of the natural sequence protein suggested that residues 25 and 78 are at the protein surface and play an important role in antiviral activity. The residue at position 45 was found to be important for maintaining the protein structure, as assessed by circular dichroism spectroscopy. The addition of Cys-Tyr-Cys resulted in a small perturbation of protein structure and a small decrease in antiviral activity.

Preparation and characterization of recombinant DNA-derived human interferon-gamma.

An attempt was made to prepare a highly purified, active recombinant DNA-derived human interferon-gamma. When the protein was denatured in urea and refolded, gel filtration and sedimentation velocity experiments indicated the presence of two forms, which are different in size and are not in a rapid reversible equilibrium. The two forms could be chromatographically separated. Far-UV circular dichroic spectra indicated the presence of secondary structures for both forms. Near-UV circular dichroic spectra revealed that the smaller form is folded into a rigid tertiary structure. The antiviral activity of the two forms of interferon-gamma showed a significant difference, i.e. the smaller form was 4-8-fold more active than the larger form. A variety of experiments show that the smaller form is more active, homogeneous, soluble, and stable than the larger form.

5'-Untranslated sequences of two structural genes in the qa gene cluster of Neurospora crassa.

The coding regions of two genes (qa-2 and qa-3) in the qa gene cluster of Neurospora crassa have been localized by nucleotide sequence analysis combined with data on previously determined NH2-terminal amino acid sequences for the proteins that these genes encode. The start point of transcription for each of these genes has been determined by nuclease S1 mapping experiments with poly(A)+RNA isolated from quinic acid-induced cultures of N. crassa. The sequences of approximately 200 nucleotides 5' to the start point of transcription have been compared with each other and with those of other eukaryotes. The results show that neither of these regions for the qa-2 nor the qa-3 genes share any significant homology with sequences apparently conserved in higher eukaryotic promoters (-25 and -70 regions). However, the qa-2 and qa-3 sequences do show homology with each other in these regions. Comparison of the 5'-flanking regions of these Neurospora genes with those of several Saccharomyces cerevisiae genes reveals a number of similarities in the region preceding the translation initiation codons.

Nucleotide sequence analysis of the chloramphenicol resistance transposon Tn9.

The transposable genetic element Tn9 consists of two direct repeats of the insertion sequence IS1 flanking a region of 1,102 base pairs which determines chloramphenicol resistance. Transposition of Tn9 leads to the duplication of a 9-base pair sequence which preexists at the site of insertion. One copy of this sequence is found at each end of the inserted element. The chloramphenicol resistance determined by Tn9, and by various other R plasmids, is due to the synthesis of the enzyme chloramphenicol acetyl transferase (CAT). This enzyme catalyses the formation of acetylated derivatives of chloramphenicol which are inactive as inhibitors of protein synthesis. By using the chain termination technique of DNA sequencing, we have now determined the nucleotide sequence of the 1,102 base pair region between the directly repeated IS1 sequence in the bacterial transposon Tn9 (encoding chloramphenicol resistance). The amino acid sequence of CAT predicted from the nucleotide sequence is identical to that determined by Shaw and coworkers. An analysis of the sequence suggests that the internal 1,102 base pair region is not directly involved in transposition.

In vivo transcription and translation of R-plasmid 538-1 DNA in Escherichia coli.

In vivo transcription and translation of R-plasmid 538-1 in E. coli was analyzed. Transcription of individual restriction fragments was determined qualitatively by utilizing the techniques developed by Southern (27), and quantitatively by carrying out DNA-RNA filter hybridization. The most active region of R-plasmid transcription in strains repressed for conjugal transfer was found to occur in the region of the R-plasmid carrying the antibiotic resistance genes. In plasmids derepressed for conjugal transfer, a high level of transcription from the transfer gene region was also observed. When strains carrying R538-1drd were induced with Hg++ a high level of transcription was observed from the region of the R-plasmid carrying the genes for Hgr. Hybrid ColE1 plasmids carrying restriction fragments from the antibiotic resistance region of R538-1 were segregated into minicells. Labeling of the minicells with 35S-methionine allowed identification of the proteins coded by the fragments. A limited number of proteins were detected, and several of these have been correlated with the antibiotic resistance genes carried by R538-1.

Transcription and translation in E. coli of hybrid plasmids containing the catabolic dehydroquinase gene from Neurospora crassa.

Two hybrid plasmids which carry the gene for Neurospora crassa catabolic dehydroquinase (C-DHQase) and complement an aroD6 (dehydroquinase-deficient) auxotroph of Escherichia coli have been analyzed. One of these contains a 2.9 kilobase (kb) fragment cloned in the HindIII site of plasmid pBR322 (pVK57) and the other contains a 6.8 kb fragment cloned in the PstI site (pVK88). Restriction enzyme mapping of these plasmids has demonstrated that the 2.9 kb fragment is totally contained within the 6.8 kb fragment. When the polarity of either the HindIII fragment or PstI fragment was reversed with respect to pBR322 no effect was observed on either the ability of the hybrid to complement an aroD- auxotroph or on the level of C-DHQase activity. In vivo transcription of plasmid pVK88 in both orientations was analyzed by RNA-DNA hybridization and by the techniques developed by Southern (1975). Approx. 40% of the plasmid-directed transcription occurred from the cloned PstI fragment and 60--70% of these N. crassa transcripts were encoded by the 2.9 kb HindIII fragment. The Southern technique allowed a further localization of the region of most extensive transcription to a 1.8 kb HindIII-EcoRI fragment. Biochemical analysis revealed that the C-DHQase protein produced by strains harboring pVK57 and pVK88 in either orientation was identical to the N. crassa enzyme. Furthermore, when these plasmids were segregated into minicells and labeled with 14C amino acids, the C-DHQase protein was synthesized at a level comparable to other plasmid-encoded proteins. Taken together, these experiments demonstrate that transcription is efficiently initiated in E. coli from a site on the cloned N. crassa DNA and that the resulting C-DHQase mRNA is efficiently and accurately translated.

Amplification in Escherichia coli of enzymes involved in genetic recombination: construction of hybrid ColE1 plasmids carrying the structural gene for exonuclease I.

Endo-R-HindIII restriction endonuclease fragments obtained from F30 and pMB9 plasmid DNAs were ligated in vitro and used to transform a recB21 recC22 sbcB15 strain of E. coli K-12. The inability of this strain to stably maintain pMB9 alone permitted the isolation of transformants that carried hybrid plasmids containing the sbcB+ allele. These transformants became sensitive to ultraviolet light and recombination defieient and showed a 25-fold increase in the level of exonuclease I activity. The stability of the sbcB hybrid plasmids and their effects on exonuclease I activity have also been determined in wild-type and recA1 genetic backgrounds. The presence of the plasmids results in a 7-fold increase in the level of exonuclease I in a wild-type strain and 15-fold increase in a recA1 strain. The increased activity in the recA1 mutant appears to be a result of increased plasmid stability in this genetic background.