Functional and molecular characterization of tumor-infiltrating lymphocytes transduced with tumor necrosis factor-alpha cDNA for the gene therapy of cancer in humans.

TNF is effective in causing the regression of selected murine tumors when administered at high concentrations. Therapeutic levels in humans cannot be obtained systemically, however, because of dose-limiting toxicity. The development of immunotherapy with IL-2 and tumor-infiltrating lymphocytes (TIL), which can accumulate at tumor sites in some patients, and of efficient retroviral techniques for gene transfer into eukaryotic cells has allowed new therapeutic approaches using TNF. We have retrovirally transduced human TIL with the gene for TNF in an attempt to deliver high concentrations of TNF to the tumor site without dose-limiting systemic toxicity. Successful gene insertion was confirmed by Southern hybridization in 16 of 16 transduced and selected TIL cultures from 15 different patients, with an estimated 28 to 93% transduced cells within each culture. Transduced selected TIL cultures produced greater amounts of TNF, compared with nontransduced controls, in 11 of 16 cultures evaluated. However, overall production of TNF was > 30-fold lower, compared with a transduced and highly selected tumor cell line control (MEL-TNF). In addition, steady state levels of vector-derived transcript in nine of 10 transduced selected TIL cultures were < 14% of the amount seen in the MEL-TNF control line. In an attempt to increase TNF production, TIL were transduced with a mutated form of TNF containing the IFN-gamma signal peptide in place of the transmembranous region, to enhance secretion into the endoplasmic reticulum. By using this vector, TNF production increased by an average of fivefold. These studies demonstrate that TIL can be genetically modified to express and secrete a protein for use in targeted cancer therapy but that partial expression blockades exist that prevent maximal cytokine production by introduced genes in TIL.

Enhanced interaction of L-selectin with the high endothelial venule ligand via selectively oxidized sialic acids.

The selectins are adhesion receptors that play key roles in leukocyte trafficking. Each has an N-terminal C-type lectin domain that binds to specific carbohydrates in a calcium-dependent manner. L-selectin recognizes sulfated, sialylated ligands on lymph node high endothelial venules. This recognition is abolished by strong periodate oxidation under conditions that destroy oligosaccharides. In contrast, mild periodate oxidation, which selectively oxidizes the side chain of sialic acid residues without affecting the underlying oligosaccharide, markedly enhances this interaction. The enhancement is calcium dependent, indicating that lectin recognition is maintained. Reduction of the sialic acid aldehydes generated by mild periodate to alcohol groups abolishes this effect. Covalent cross-linking of the oxidized ligand to L-selectin can be demonstrated, suggesting Schiff base formation between lysine residues of the selectin and the newly formed aldehydes. Such selectively oxidized sialylated ligands could be used to probe the lectin domains of the selectins and to identify lysine residues near the binding site. Also, this approach could be used to design drugs for disrupting leukocyte-endothelial interactions leading to pathological inflammation.

In vivo activity of tumor necrosis factor (TNF) mutants. Secretory but not membrane-bound TNF mediates the regression of retrovirally transduced murine tumor.

We have previously demonstrated that murine tumor cells transduced with a retrovirus containing the cDNA encoding wild-type human TNF regress in vivo when injected into immunocompetent mice; this regression is T cell mediated. To determine whether membrane-associated or secreted TNF was responsible for tumor regression, we transduced a cloned murine fibrosarcoma 205 F4 with retroviruses encoding modified human TNF genes. The cloned tumor lines of one retroviral transduction expressed only membrane bound 26-kDa TNF. This TNF could not be cleaved or secreted, but was present on the cell surface. A second retrovirus caused the expression of only secretory 17-kDa TNF, as the transmembrane domain of the cDNA was deleted. The TNF produced by tumor cells transduced with either retroviral vector was functional in vitro as direct lysis of the TNF-sensitive target L929 by transduced tumor cells was demonstrated. The TNF present on 26-kDa expressing tumors was membrane bound as supernatants from cultured 17-kDa TNF expressing tumor cells but not 26-kDa TNF expressing tumors mediated the lysis of L929 cells. Both tumors were injected s.c. into syngeneic mice and tumor growth was measured serially. In repeated experiments, 26-kDa TNF expressing tumors grew progressively in all mice. In contrast, 17-kDa TNF expressing tumors grew for 10 days and then regressed with all animals free of tumor at 28 days. Tumor regression was abrogated by in vivo injection of an anti-TNF antibody. Similar results were obtained in a second tumor model, 203 E4. Thus regression of TNF transduced tumors in vivo requires secretion of TNF, as membrane-bound TNF is insufficient to elicit the host response.

Multidrug resistance after retroviral transfer of the human MDR1 gene correlates with P-glycoprotein density in the plasma membrane and is not affected by cytotoxic selection.

Multidrug resistance (MDR) in mammalian cells is associated with the expression of the MDR1 gene encoding P-glycoprotein (P-gp), an and active efflux pump for various lipophilic compounds. MDR transfectants can be isolated after MDR1 gene transfer and selection with cytotoxic drugs; low levels of drug resistance have also been observed in unselected NIH 3T3 mouse cells after retrovirus-mediated transfer of mouse mdr1 cDNA. MDR cell lines possess multiple phenotypic changes, suggesting that P-gp function could be complemented by some additional mechanisms associated with cytotoxic selection. To determine whether cytotoxic selection contributes to the MDR phenotype of MDR1-expressing cells, NIH 3T3 cells infected with a recombinant retrovirus carrying the human MDR1 gene were selected by two different procedures: (i) noncytotoxic selection for increased P-gp expression on the cell surface by multiple rounds of immunofluorescence labeling and flow sorting or (ii) one or more steps of selection with a cytotoxic drug. The levels of MDR in both types of infectants showed an excellent correlation with the P-gp density in the plasma membrane, expressed as immunoreactivity with a P-gp-specific antibody normalized by reactivity with an antibody against an unrelated antigen. Cytotoxic selection conferred no additional increase in resistance relative to P-gp density. These results indicate that P-gp density in the plasma membrane may be sufficient to determine the level of MDR.

Murine tumor cells transduced with the gene for tumor necrosis factor-alpha. Evidence for paracrine immune effects of tumor necrosis factor against tumors.

Studies of the anti-tumor activity of TNF-alpha in vivo have been hampered by the need to administer systemically toxic doses of the cytokine to obtain a curative response. To facilitate studies of the effect of high local concentrations of TNF-alpha on tumor growth and host immunity, a newly induced murine sarcoma was transduced with the gene for human TNF-alpha and the biologic characteristics of these cells were examined. We identified high and low TNF-producing tumor clones which exhibited stable TNF secretion over time. Significant amounts of membrane associated TNF were found in a high-TNF producing clone as well. No difference in the in vitro growth rates between TNF-producing and nonproducing cell lines was observed. In contrast, in vivo studies demonstrate that although unmodified parental tumor cells grew progressively when implanted s.c. in animals, tumor cells transduced with the TNF gene were found to regress in a significant number of animals after an initial phase of growth. This effect correlated with the amount of TNF produced and could be blocked with a specific anti-TNF antibody. Regressions of TNF-producing cells occurred in the absence of any demonstrable toxicity in the animals bearing these tumors. TNF-producing tumor cells could function in a paracrine fashion by inhibiting the growth of unmodified, parental tumor cells implanted at the same site. The ability of tumor cells to regress was abrogated by in vivo depletion of CD4+ or CD8+ T cell subsets and animals that had experienced regression of TNF-producing tumors rejected subsequent challenges of parental tumor. Our studies thus show that tumor cells elaborating high local concentrations of TNF regress in the absence of toxicity in the host and that this process requires the existence of intact host immunity. Studies of the lymphocytes infiltrating the gene modified tumors and attempts to use TNF gene modified tumor infiltrating lymphocytes to deliver high local concentrations of TNF to the tumor site without inducing systemic toxicity are underway.

A nonsecretable cell surface mutant of tumor necrosis factor (TNF) kills by cell-to-cell contact.

In addition to the induction of tumor regression, tumor necrosis factor (TNF) has been implicated as the causative agent in a number of pathologies, including cachexia, septic shock, rheumatoid arthritis, autoimmunity, and induction of HIV expression. We propose that this complex physiology might be manifest by different forms of TNF: the 17 kd secretory component, the 26 kd transmembrane form, or both. To determine whether the 26 kd form of TNF was biologically active and whether its biology differed from that of the secretory component, we generated uncleavable and solely secretable mutants of TNF and studied their biological activities. We found that an uncleavable mutant of the 26 kd cell surface transmembrane form of TNF kills tumor cells and virus-infected cells by cell-to-cell contact, and that TNF need not be internalized by its target to kill. Thus, the 26 kd integral transmembrane form of TNF may function in vivo to kill tumor cells and other targets locally in contrast to the systemic bioactivity of the secretory component.

Laryngeal carcinoma in a Lynch syndrome II kindred.

Hereditary nonpolyposis colorectal cancer (HNPCC) accounts for about 4% to 6% of the total colorectal cancer burden. It is subdivided into Lynch syndrome I and II. Lynch syndrome I is characterized by an autosomal dominant inheritance pattern for site-specific, early onset, adenocarcinoma of the colon, with proximal predominance and an excess of synchronous and metachronous colonic cancers. Lynch syndrome II (cancer family syndrome) shows these same colon cancer characteristics, but differs in that there is an excess proclivity of other forms of cancer, particularly of the endometrium and ovary. This article documents a family that shows features of Lynch syndrome II. Unique aspects pertain to a patient who is in the direct genetic lineage (whose five brothers manifested colonic cancer), but who developed carcinoma of the uterine cervix at age 34 and laryngeal cancer at 60. The pedigree also shows uterine cervical carcinoma among other patients at genetic risk. Her son, who is a nonsmoker and nondrinker, manifested laryngeal cancer at age 31. These observations appear to add new information about tumor heterogeneity in HNPCC.

Transformation of human mammary epithelial cells by oncogenic retroviruses.

We have introduced viral oncogenes into human mammary epithelial cells through the use of murine retroviruses. A continuous cell line (184A1N4) derived from benzo(a)pyrene treatment of normal breast epithelial cells was used as a recipient for the ras, mos, and T-antigen oncogenes. Each of these oncogenes enabled the 184A1N4 cells to grow in a selective medium, thus demonstrating the potential utility of these cells for oncogene detection and isolation. 184A1N4 cells transformed by T-antigen were nontumorigenic in athymic mice, but v-ras transformants were weakly tumorigenic. Transformants bearing both the T-antigen and ras oncogenes were strongly tumorigenic, however. The karyotype of these double transformants shows a high degree of stability. These results demonstrate the stepwise acquisition of the fully malignant phenotype by normal human epithelial cells in vitro.

Natural history of colorectal cancer in hereditary nonpolyposis colorectal cancer (Lynch syndromes I and II).

Approximately 5 to 6 percent of the total colorectal cancer burden is accounted for by hereditary nonpolyposis colorectal cancer (HNPCC). Because clinical premonitory signs such as those seen in familial polyposis coli (FPC) are lacking, the clinician must recognize clinical findings and family history typical of HNPCC. The authors have described colorectal cancer expression from a survey of ten HNPCC kindreds. Kindred members with colorectal cancer differed significantly (P less than .05) from patients with sporadic colorectal cancer: 1) mean age of initial colon cancer diagnosis was 44.6 years; 2) 72.3 percent of first colon cancers were located in the right colon, and only 25 percent were in the sigmoid colon and rectum; 3) 18.1 percent had synchronous colon cancers; and 4) 24.2 percent developed metachronous colon cancer, with a risk for metachronous lesions in ten years of 40 percent. Affecteds and their first-degree relatives should undergo early intensive education and surveillance. In families with an early age of onset, colonoscopy should begin at age 25, and biannually thereafter, with fecal occult blood testing of the stool semiannually. Third-party carriers must become more responsive to the costly surveillance measures required for these otherwise healthy patients.

An altered pattern of cross-resistance in multidrug-resistant human cells results from spontaneous mutations in the mdr1 (P-glycoprotein) gene.

Multidrug resistance in human cells results from increased expression of the mdr1 (P-glycoprotein) gene. Although the same gene is activated in cells selected with different drugs, multidrug-resistant cell lines can be preferentially resistant to their selecting agent. The mdr1 cDNA sequence from vinblastine-selected KB cells, which are uniformly resistant to different lipophilic drugs, was compared with the corresponding sequence from colchicine-selected KB cells preferentially resistant to colchicine. These sequences differ at three positions, resulting in a single amino acid change in P-glycoprotein. These differences result from mutations that occurred during colchicine selection. The appearance of these mutations coincides with the emergence of preferential resistance to colchicine. We have constructed biologically active mdr1 cDNA clones that express either wild-type or mutant P-glycoprotein. Multi-drug-resistant transfectants obtained with the mutant sequence were characterized by increased relative resistance to colchicine compared with transfectants obtained with wild-type sequence. mdr1 mutations are therefore responsible for preferential resistance to colchicine in multidrug-resistant KB cells.

Differential diagnosis of hereditary nonpolyposis colorectal cancer (Lynch syndrome I and Lynch syndrome II).

Increasing recognition of the statistical burden posed by HNPCC (5 to 6 percent of all colorectal cancer) mandates that physicians have a better understanding of the genetics, natural history, and distinction between the hereditary site-specific variant (Lynch syndrome I) and the Cancer Family Syndrome (Lynch syndrome II). The authors report detailed cancer (all sites) family histories on two prototype families with Lynch syndrome I (Family R) and Lynch syndrome II (Family N), which have been under investigation for more than two decades. Emphasis is placed on shared clinicogenetic features; namely, early age of onset of colonic cancer (approximately age 44), multiple primary colonic cancer (24 percent of cases showed metachronous colonic cancer), predominance of proximal colonic cancer location (approximately 65 percent in the proximal colon), and vertical transmission consonant with an autosomal dominantly inherited factor. An increased predilection for extracolonic cancer, particularly endometrial carcinoma, occurs in Lynch syndrome II and is the primary basis for distinction from Lynch syndrome I. Surveillance and management programs must be wholly responsive to these natural history features.

Flat adenomas in a colon cancer-prone kindred.

We describe new pathologic findings in a hereditary nonpolyposis colorectal cancer family. Affected family members developed multiple small adenomas with right-sided predominance; many adenomas had an unusual appearance featuring slightly elevated lesions with adenomatous changes confined to the upper regions of the colonic crypts. We have adopted the previously established term "flat adenoma" for these lesions. This phenotype may be a morphologic marker for at least one subset of hereditary nonpolyposis colorectal cancer.

A novel form of TNF/cachectin is a cell surface cytotoxic transmembrane protein: ramifications for the complex physiology of TNF.

Tumor necrosis factor (TNF) is a monocyte-derived cytotoxin that has been implicated in tumor regression, septic shock, and cachexia. The mechanism by which TNF induces these different disease states is unclear. We have identified and characterized a novel, rapidly inducible cell surface cytotoxic integral transmembrane form of TNF. The existence and behavior of this novel form of TNF may explain the complex physiology of this molecule. We suggest that activated monocytes synthesize transmembrane TNF at the site of inflammation and kill their targets by either cell-to-cell contact or local release of the TNF secretory component. In contrast, septic shock and cachexia may result from either acute or chronic systemic activation of monocytes, resulting in the widespread release of TNF secretory component into the circulation of the affected individual. We further suggest that cell borne cytokines and cytotoxins may be the primary mediators of directed inflammatory responses.

Recognition and treatment of patients with hereditary nonpolyposis colon cancer (Lynch syndromes I and II).

Primary genetic factors are etiologic in at least 5-10% of patients with colon cancer. The polyposis syndromes (FPC) are easily identified examples because of the spectacular number of polyps. The hereditary nonpolyposis syndromes (HNPCC), although five times more common than FPC, are usually not recognized because they do not have such a distinctive clinical, premonitory genetic marker. Colorectal cancer expression was surveyed in 10 extended, thoroughly documented HNPCC kindreds. One hundred sixteen patients were found to have 183 colorectal cancers. Despite the striking family history, less than 5% were correctly treated by subtotal colectomy. This provided a unique opportunity to study the natural history. Five findings differed significantly (p less than 0.05) from patients with sporadic colon cancer: (1) mean age of initial colon cancer diagnosed was 45.6 years; (2) 69.1% of first colon cancers were located proximal to the splenic flexure of the colon; (3) 18.1% had synchronous colon cancer; (4) 24.2% had metachronous colon cancer develop with life table analysis showing the risk for a metachronous lesion at 10 years to be 40%; and (5) only 23.3% of cancers were located in the sigmoid colon or rectum. Based on this data, it is recommended that the family history of all patients with a newly diagnosed colon cancer be evaluated for evidence of this syndrome. If an autosomal dominant inheritance pattern emerges, an in-depth genetic investigation is indicated. When HNPCC is confirmed, the following recommendations apply: a subtotal abdominal colectomy is indicated at the time of the initial colon cancer because of the risk of synchronous and metachronous lesions. The rectum should be spared in favor of careful lifetime surveillance because of the proclivity for proximal colon cancer involvement. As yet unaffected members of a newly diagnosed HNPCC kindred who are in the "direct genetic line" should be cautioned that they are at 50% risk and must begin an intensive surveillance program beginning in the third decade with careful attention to the right colon. Patients from newly diagnosed HNPCC families who have had a previous conventional colectomy for colon cancer should, at the very least, enter an intensive surveillance program; a prophylactic completion subtotal colectomy should be considered for patients who are less than totally compliant.

Transformation mediated by the SV40 T antigens: separation of the overlapping SV40 early genes with a retroviral vector.

A murine retroviral vector has been used to separate physically the overlapping genes encoded by SV40. This minimal retroviral vector contains LTRs and other cis-acting signals required for infectious RNA virus propagation. We placed the SV40 early region within this DNA and after transfection of cells producing helper Moloney murine leukemia virus, SV40 retroviruses (MV40) could be rescued. Cytoplasmic spliced large T and small t transcripts, as well as unspliced transcripts, are packaged into virions with equal efficiency. Pure SV40 large T retroviruses can be cloned from these heterogeneous virus stocks by secondary transformation of rodent cells. The large T retrovirus stocks morphologically transform primary or established mouse and rat lines with high efficiency. There is little difference in transformation either by agar assay or focus formation between retroviruses carrying both SV40 genes or large T alone. We present quantitative data that demonstrate that abortive transformation of rodent cells by SV40, transient expression of the transformed phenotype after infection, is not manifested by MV40. Thus abortive transformation is not the result of a weakly dominant transforming gene, but rather of the normally inefficient mode of integration and early gene expression of SV40 upon infection of rodent cells.

Enhanced transformation by a simian virus 40 recombinant virus containing a Harvey murine sarcoma virus long terminal repeat.

We have constructed a recombinant simian virus 40 (SV40) DNA containing a copy of the Harvey murine sarcoma virus long terminal repeat (LTR). This recombinant viral DNA was converted into an infectious SV40 virus particle and subsequently infected into NIH 3T3 cells (either uninfected or previously infected with Moloney leukemia virus). We found that this hybrid virus, SVLTR1, transforms cells with 10 to 20 times the efficiency of SV40 wild type. Southern blot analysis of these transformed cell genomic DNAs revealed that simple integration of the viral DNA within the retrovirus LTR cannot account for the enhanced transformation of the recombinant virus. A restriction fragment derived from the SVLTR-1 virus which contains an intact LTR was readily identified in a majority of the transformed cell DNAs. These results suggest that the LTR fragment which contains the attachment sites and flanking sequences for the proviral DNA duplex may be insufficient by itself to facilitate correct retrovirus integration and that some other functional element of the LTR is responsible for the increased transformation potential of this virus. We have found that a complete copy of the Harvey murine sarcoma virus LTR linked to well-defined structural genes lacking their own promoters (SV40 early region, thymidine kinase, and G418 resistance) can be effectively used to promote marker gene expression. To determine which element of the LTR served to enhance the biological activity of the recombinant virus described above, we deleted DNA sequences essential for promoter activity within the LTR. SV40 virus stocks reconstructed with this mutated copy of the Harvey murine sarcoma virus LTR still transform mouse cells at an enhanced frequency. We speculate that when the LTR is placed more than 1.5 kilobases from the SV40 early promoter, the cis-acting enhancer element within the LTR can increase the ability of the SV40 promoter to effectively operate when integrated in a murine chromosome. These data are discussed in terms of the apparent cell specificity of viral enhancer elements.

Cysteine auxotrophy of human leukemic lymphoblasts is associated with decreased amounts of intracellular cystathionase protein.

A series of human lymphoblastoid cell lines derived from nonleukemic donors are known to be cysteine prototrophs (cys+), while several lymphoblastoid lines derived from leukemic donors are cysteine auxotrophs (cys-). We have tested representative cell lines of each type for their content of cystathionase enzyme activity by a specific catalytic assay and their total cystathionase protein content by immunoprecipitation of in vivo labeled protein. There was a close correlation between the cellular content of the enzyme as determined in the two assays. Specifically, those cys+ lines having readily measureable enzyme by catalytic assay were found to contain significantly higher levels of immunoprecipitable Mr 43 000 cystathionase subunit than those cys- lines tested which were depleted in active enzyme. Thus, the absolute cysteine requirement of the leukemic, cys- cell lines tested is likely due to an intracellular reduction of cystathionase protein.

Cysteine auxotrophy of human leukemic lymphoblasts is associated with decreased amounts of intracellular cystathionase messenger ribonucleic acid.

A series of human lymphoblastoid cell lines derived from nonleukemic donors are known to be cysteine prototrophs (cys+), while several lymphoblastoid lines derived from leukemic donors are cysteine auxotrophs (cys-). Compared to the cys+ lines, the cysteine auxotrophic cell lines have been found to contain reduced amounts of cystathionase, the enzyme which catalyzes te conversion of cystathionine to cysteine, the terminal step in the cysteine biosynthetic pathway. We have developed a semiquantitative in vitro translation assay for the messenger ribonucleic acid (mRNA) which encodes cystathionase and have tested two pairs of cys-, cystathionase- and cys+, cystathionase+ cell lines for their intracellular content of this RNA. There is a close correlation between the cys- phenotype, reduced intracellular cystathionase content, and the intracellular content of enzyme-specific messenger RNA. Thus, the absolute cysteine requirement of the human leukemic, cys-, cystathionase- cell lines tested is likely due to an intracellular reduction in the functional messenger RNA coding for cystathionase. Among the possibilities that could amount for these phenomena are (1) a reduction in the rate of transcription of the cystathionase mRNA precursor, (2) an alteration in the nuclear and/or cytoplasmic processing of the primary transcripts, or (3) an alteration in the mRNA structure, resulting in a decreased efficiency of translation.