Effect of Corynebacterium parvum on the immune response in guinea pigs. II. Passive transfer of enhanced anamnestic response and delayed hypersensitivity observed after treatment with Corynebacterium parvum]

After intradermal immunization with a mixture of Corynebacterium parvum (C. parvum) and ovalbumin guinea pigs show a markedly increased anamnestic response to an intradermal booster of ovalbumin as compared to controls treated with ovalbumin only. At the same time a reaction of delayed type hypersensitivity is observed in the treated animals, but not in controls. The enhanced anamnestic response as well as the posivitive skin reaction were transferred to strain 2 histocompatible guinea pigs by peripheral blood leukocytes as well as by peritoneal exudate cells. Passive transfer was not obtained after prior irradiation of donor animals.     

Corynebacterium parvum as an adjuvant for Trypanosoma cruzi epimastigote vaccines: a comparison with saponin and Bordetella pertussis.

The effect was compared in CBA mice of adding Corynebacterium parvum, saponin, and Bordetella pertussis to living or killed Trypanosoma cruzi (Y strain) epimastigote vaccines on the induction of protective immunity against subcutaneous (s.c.) challenge with blood trypomastigotes. The addition of C. parvum to a low dose of T. cruzi vaccine, which alone was non-protective, generated a greater degree of protection than did saponin or B. pertussis. C. parvum alone increased resistance to infection to a variable and usually weak extent. The addition of C. parvum to larger doses of T. cruzi vaccine, which were themselves sufficient to elicit some degree of protection, improved resistance when the challenge was given 1 or 12 weeks after immunization, but lowered it at 3 weeks. It is concluded that the comparative efficacy of adjuvants for T. cruzi vaccines needs to be assessed on 3 parameters: (1) the dose of antigen, (2) the dose of adjuvant and (3) the time interval between immunization and challenge.     

Inhibition of cell-mediated cytotoxicity to tumor alloantigens by systemic administration of Corynebacterium parvum

Summary Intravenous administration of C. parvum following SC immunization with a tumor allograft markedly impaired the generation and expression of primary cell-mediated cytotoxicity to the immunizing alloantigens. For the inhibition to become manifest, a minimum of 4 days of in vivo exposure to C. parvum was necessary, and the effect was seen only when C. parvum was injected during the developing immune response; injection prior to or simultaneously with alloimmunization resulted in no impairment. The inhibition of cell-mediated cytotoxicity was selective. While spleen cell populations exhibited decreased cytotoxic activity, lymphoid cell populations from lymph nodes and blood were unaffected. The inhibitory effect was restricted to the cellular component of the immune response. Complement-dependent cytotoxic antibodies developed normally and their titer was comparable to that of control alloimmunized mice.     

Comparison of the host ranges and antigenicity of Cryptosporidium parvum and Cryptosporidium wrairi from guinea pigs.

Oocysts of a Cryptosporidium isolate from guinea pigs were not infectious for adult mice, but were infectious for two of three newborn calves and for suckling mice. However, oocysts isolated from calves or mice infected with guinea pig Cryptosporidium were not infectious for guinea pigs. Four isolates of C. parvum from calves were incapable of infecting weanling guinea pigs. Microscopic examination of tissue from the colon and cecum of suckling guinea pigs inoculated with C. parvum revealed sparse infection of some pups. These host range studies and previously described differences in 125I-labeled oocyst surface protein profiles between Cryptosporidium sp. from guinea pigs and C. parvum suggest they are distinct species. We propose the name Cryptosporidium wrairi be retained. Studies with monoclonal antibodies indicate that C. wrairi and C. parvum are antigenically related.     

Effects of Corynebacterium parvum on depressed immunological reactions in EL-4-tumor-bearing mice

Summary Effects of Corynebacterium parvum on the development of plaque-forming cells (PFC), cell-mediated cytotoxicity (CMC), and delayed footpad reaction (DFR) to chicken erythrocytes (CRBC) were investigated in EL-4-bearing syngeneic mice. PFC, CMC, and DFR responses after the primary immunization were suppressed in tumor-bearing mice and restored by C. parvum treatment. PFC and CMC responses in tumor-bearing mice were restored by the transfer of spleen cells of C. parvum-treated normal mice. Such powers of recovery were abrogated by the removal of glass-adherent cells but not by the removal of -positive or Ig-positive cells. DFR was suppressed not only in the primary but also in the secondary immunization, in contrast to PFC and CMC; the secondary responses of these types were not suppressed in tumor-bearing mice. Positive DFR was not elicited in tumor-bearing mice after adoptive transfer of sensitized lymphocytes from normal immune donors. The DFR became positive in such tumor-bearing recipients when they were treated by C. parvum. Macrophage functions in the induction phase of the immune response as accessory cells and in the expression of DFR as secondary cells appear to be suppressed in tumor-bearing mice and restored by C. parvum.     

Inhibition of cell-mediated cytotoxicity by Corynebacterium parvum

Summary We investigated the effect of altering dose and route of Corynebacterium parvum (C. parvum) administration on the adjuvant's inhibition of cell-mediated cytotoxicity (CMC). Primary in vivo and secondary in vitro CMC of C57B1/6 mice alloimmunized to P815 were depressed if C. parvum was administered systemically (IV or IP) but not when it was given SC. Similarly, only systemic C. parvum generated cells capable of suppressing in vitro CMC. Primary and secondary CMC in spleen was equally inhibited by 700 and 70 g, whereas suppressor cell activity was marked with 700 g and minimal with 70 g. Administration of C. parvum SC admixed with alloantigen resulted in early enhancement and late depression of primary CMC. Secondary CMC was depressed but suppressor activity was absent. Dissociation of CMC depression from suppressor cell generation indicates that these phenomena can be separated under certain conditions.     

Protection by polyI:polyC against infection with herpes simplex virus in mice pretreated with Corynebacterium parvum

The effect of Corynebacterium parvum on the protection by polyinosinic-polycytidylic acid (polyI:polyC) against lethal infection with Herpes simplex virus type 1 (HSV) was studied in mice. Pretreatment with C. parvum resulted in prolonged survival times in all experiments. One third of the mice survived an infection with 100 LD50, whereas all mice died when treated with polyI:polyC alone. Increased protection was observed up to 6 weeks after pretreatment and only seen when both C. parvum and polyI:polyC were given at the same site of injection (intraperitoneally). Protection against HSV correlated with increased interferon (IFN) activities induced by polyI:polyC in the peritoneal cavity of C. parvum-pretreated mice. In these mice, natural killer cell activity of peritoneal exudate cells (PEC) was also augmented in response to polyI:polyC. Protection was markedly decreased by intraperitoneal injection of silica or of an antiserum against murine IFN. It appears that increased local levels of IFN presumably produced by macrophages in response to polyI:polyC in C. parvum-pretreated mice play the major role in the antiviral defence in our model and that activation of NK cells may be a secondary effect of IFN.     

Comparison of the Host Ranges and Antigenicity of Cryptosporidium parvum and Cryptosporidium wrairi from Guinea Pigs

ABSTRACT. Oocysts of a Cryptosporidium isolate from guinea pigs were not infectious for adult mice, but were infectious for two of three newborn calves and for suckling mice. However, oocysts isolated from calves or mice infected with guinea pig Cryptosporidium were not infectious for guinea pigs. Four isolates of C. parvum from calves were incapable of infecting weanling guinea pigs. Microscopic examination of tissue from the colon and cecum of suckling guinea pigs inoculated with C. parvum revealed sparse infection of some pups. These host range studies and previously described differences in ¹²⁵I-labeled oocyst surface protein profiles between Cryptosporidium sp. from guinea pigs and C. parvum suggest they are distinct species. We propose the name Cryptosporidium wrairi be retained. Studies with monoclonal antibodies indicate that C. wrairi and C. parvum are antigenically related.     

Evidence for granulocyte-mediated macrophage activation after C. parvum immunization

It has been previously demonstrated that at the peak of the peritoneal response to Corynebacterium parvum (Day 4), cytolytic macrophages can be characterized by the presence of intracellular bacteria. In the present study, the role of neutrophils in the activation of peritoneal macrophages by C. parvum was investigated. Inflammatory neutrophils isolated 5 hr after ip administration of C. parvum were transferred to normal, syngeneic mice and the peritoneal macrophages of recipients harvested 4 days later were tested for cytoxicity against HeLa cells. Neutrophils isolated from mice 5 hr after C. parvum immunization were effective in inducing cytolytic macrophages. Less than 100-fold as much bacteria was needed to induce comparable levels of cytotoxic activity when introduced inside granulocytes. Neutrophils obtained from mice 48 hr after C. parvum injection or mononuclear cells were not good macrophage activators. Viable neutrophils were not required as freeze-thawed cells were able to activate macrophages in recipient mice. The intracellular distribution of C. parvum changed dramatically with time. Initially almost all bacteria were found within neutrophils. By 24 hr, many macrophages contained either bacteria or granulocytes which had ingested C. parvum. Pyridine extracts of C. parvum, which do not activate peritoneal macrophages when injected directly into mice, did not induce neutrophils capable of activating macrophages. The residue of pyridine-extracted C. parvum did induce neutrophils that could activate macrophages when transferred. The results suggest that processing of the bacteria by inflammatory granulocytes may be an obligatory step in macrophage activation by this agent. The peak response occurred earlier than T-cell immunity is usually observed and it is suggested that direct activation of macrophages via ingestion of neutrophils may represent the earliest stage of macrophage activation by C. parvum.     

A comparative study on the biology of Cryptosporidium sp. from guinea pigs and Cryptosporidium parvum (Apicomplexa).

Cryptosporidium sp. from guinea pigs and C. parvum were compared morphologically, electrophoretically, and for the ability to infect suckling mice. Oocysts from guinea pigs measured 5.4 x 4.6 (4.8-5.6 x 4.0-5.0) microns and had a shape index (length/width) of 1.17 (1.04-1.33). Oocysts of C. parvum were similar and measured 5.2 x 4.6 (4.8-5.6 x 4.2-4.8) microns with a shape index of 1.16 (1.04-1.33). All suckling mice inoculated with oocysts of C. parvum became infected, whereas most, but not all, mice fed oocysts of the guinea pig isolate also became infected. However, mice inoculated with oocysts from guinea pigs produced on average 100-fold fewer oocysts by day 7 postinoculation than did mice infected with C. parvum, and the resulting infections were sparse and patchy along the ileum. Electrophoretic profiles were similar, but 125I surface labeling of outer oocyst wall proteins revealed striking differences between the two isolates. Cryptosporidium parvum had a wide molecular size range of 125I-labeled bands, whereas C. sp. from guinea pigs had a banding pattern clustered between 39 and 66 kDa, with a smaller number of bands greater than 100 kDa.     

Immunotherapy of lung cancer with Corynebacterium parvum

Summary Thirty-one patients with inoperable carcinoma of the lung, excluding oat-cell carcinoma, were randomized to receive either chemotherapy alone, with methyl CCNU and vinblastine every 6–8 weeks (15 Pts) or such chemotherapy plus immunotherapy with IV infusions of Corynebacterium parvum (16 Pts). Prior duration of the disease was longer, and more patients had received previous therapy, in the immunotherapy group; these groups were otherwise very similar. In vitro lymphocyte response to phytohemagglutinin did not change significantly in either group, but the weaker response to Varidase declined in both groups after chemotherapy. An increased baseline level of circulating B lymphocytes was sharply reduced in the C. parvum group. There were no differences in -globulins or delayed skin test responses between immunotherapy and control patients at entry into this study or on follow-up. Median survival from entry was longer in the immunotherapy group (6 months) than in the control group (3 months), but this difference was not statistically significant and only two patients in each group lived for more than 11 months. It is conceivable that more benefit from C. parvum might have been recorded had more effective chemotherapy been available.     

Therapeutic application of various immunostimulators on the L2C guinea-pig leukemia

Summary Immunostimulators such as Corynebacterium parvum (C. parvum), Bacillus Calmette-Guerin (BCG), pyran copolymer, and glucan were examined in the guinea pig L ₂ C lymphoblastic leukemia model to determine their capacity for therapeutic modulation of the immune response of the host toward controlling leukemic cell proliferation. The dose, route, and frequency of administration of the stimulators were also evaluated as a function of time in order to obtain an optimal antileukemic effect. Results indicated that only C. parvum and BCG were capable of significantly increasing host survival when given 1 day after an inoculation of 1.5×10 ⁴ viable leukemic cells. Administration of BCG or C. parvum, alone or in combination with irradiated blast cells on either days 4 or 7, was totally ineffective in prolonging survival. In the majority of cases, enhanced leukemic growth was observed on these days. The combination of BCG and/or C. parvum with irradiated syngeneic blast cells given 24 h after leukemia inoculation promoted a synergistic response with a significant increase in median survival time and a number of long-term survivors.This work was supported by contract N01-CP-53566 within the Virus Cancer Program of the National Cancer Institute     

Failure of Corynebacterium parvum to inhibit antipyrine metabolism in man

Summary In animals, Corynebacterium parvum lowers the rate of drug metabolism and enhances the pharmacologic effect of drugs requiring hepatic microsomal enzyme activity for elimination. A pilot study was conducted to assess this drug interaction in patients given clinical protocol doses of C. parvum. In individual patients, C. parvum did not reduce microsomal drug metabolism as measured by antipyrine half-life. Conversely, antipyrine elimination appeared to be enhanced in 10 of 14 patients. Results from this small heterogenous patient group are not definitive, and further studies are needed to determine the clinical significance of the effects of nonspecific immunotherapy on drug metabolism.     

Role of Corynebacterium parvum in the activation of peritoneal macrophages. II. Identification of distinguishable anti-tumor activities by macrophage subpopulations

Two different mechanisms of murine macrophage (MP) antitumor activity are described in this report. C. parvum-activated peritoneal MPs were tested for cytotoxic and cytostatic activity 4 days after ip immunization. Cytotoxic activity could be distinguished from cytostatic activity using two different assay protocols. When MPs were separated by 1g velocity sedimentation, cytotoxic MPs were confined to high velocity fractions. In contrast, cytostatic MPs were found in cell fractions with velocities as low as 5.2 mm/hr. These two MP activities were also distinguishable by culturing at 37 degrees C for 24 hr. Cytotoxicity was abrogated when MPs were incubated in MEM, or MEM supplemented with lymphokine (LK) or indomethacin. In contrast, cytostasis remained at high levels when the cells were incubated with LK or indomethacin. Cytotoxicity was not retained after overnight culture even if LPS was present, or if various spleen or non-adherent peritoneal exudate cells were cocultured with the cytotoxic effector cells. Assays done to determine the presence of suppressor cells failed to find any inhibitory cell type. The phagocytic index, acid phosphatase activity, and H2O2 secretion were also measured before and after overnight culture. Acid phosphatase and phagocytic activities did not decline whereas H2O2 secretion declined significantly. These data indicate that in response to C. parvum, at least two different effector cell types with distinct antitumor activities are generated. Cytotoxicity, like the ability of cells to secrete H2O2, is found to be a short-lived function of CP stimulated MPs. In contrast, cytostasis is a function retained longer by MPs in culture.     

Biological activity of leukotriene B4 analogs: inhibition of guinea pig eosinophil migration in vitro by the 2,6-disubstituted pyridine analogs U-75,302 and U-75,485.

A "late phase" antigen-induced bronchoalveolar eosinophilia has been demonstrated in ovalbumin sensitized guinea pigs (1,2). This in vivo response to antigen inhalation can be inhibited by a 2,6-disubstituted pyridine analog of LTB4, U-75,302(2) (3). In the present study, the mechanism of the drug action was studied by assessing the activity of U-75,302 and a second analog, U-75,485 to displace [3H]-leukotriene B4 binding at the guinea pig eosinophil membrane, as well as their action as chemoattractants or inhibitors of the directional migration of guinea pig eosinophils in vitro. Radioligand competition experiments demonstrated that both analogs interacted strongly with the high affinity LTB4 binding sites on guinea pig eosinophil membrane. Both analogs are powerful chemoattractants for guinea pig eosinophils since they induced directional migration of guinea pig eosinophils when administered alone. In addition, when the cells were treated with either analog and their chemotaxis response was measured in response to a natural chemoattractant, both U-75,302 and U-75,485 at concentrations of 0.1 to 100 microM dose dependently inhibited the LTB4 induced chemotaxis response. The EC50s obtained for U-75,302 and U-75,485 as inhibitors of LTB4 induced guinea pig eosinophil chemotaxis were estimated to be 11.5 +/- 5.5 microM and 5.4 +/- 2.5 microM respectively. Under the same conditions, they had no significant effect upon eosinophil migration induced by zymosan activated plasma at concentrations below 100 microM. We suggest that the inhibition of antigen-induced eosinophil infiltration in guinea pig airway in vivo by U-75,302 or U-75,485 may be a result of partial antagonism or desensitization at the LTB4 receptor level of guinea pig eosinophils.     

C1s-induced vascular permeability in C2-deficient guinea pigs

Normal guinea pigs that have been intradermally injected with C1s exhibit increased vascular permeability at the injection site. Guinea pigs that are genetically deficient in complement component C2 do not exhibit increased vascular permeability when given a similar injection. The C2-deficient guinea pigs respond normally to injections of bradykinin and kallikrein, suggesting that these animals can respond to kinins and have a normal kininogen pathway. When the C2-deficient guinea pigs are given guinea pig C2 before C1s injection, increased vascular permeability is observed. These results demonstrate a definite requirement for complement component C2 in the generation of C1s-induced vascular permeability.     

Disseminated intravascular coagulation following intravenous Corynebacterium parvum injection in the rat

Summary Corynebacterium parvum (C. parvum) was administered by a single IV injection (14 mg/kg) to rats, and platelet counts, plasma fibrinogen concentrations and thrombin clotting times were monitored for up to 7 weeks. During this time histological and ultrastructural studies were also conducted. Thrombocytopoenia, hypofibrinogenaemia, and prolongation of the thrombin clotting time rapidly followed C. parvum injection and were accompanied by the appearance of platelet clumps and fibrin within blood vessels in a variety of tissues. This initial episode of disseminated intravascular coagulation (DIC) subsided 12–24 h after injection, but a more prolonged second episode of DIC occurred 1–3 days after injection. The results suggest that caution should be observed when systemic immunotherapy with C. parvum is proposed.     

Antibodies to guinea pig lymphokines. II. Suppression of delayed hypersensitivity reactions by a "second generation" goat antibody against guinea pig lymphokines.

A "second generation" antibody to a highly purified lymphocyte product was raised in a goat against material eluted from a rabbit anti-guinea pig lymphokine immunoadsorbent column. This anti-lymphokine serum, in constrast to anti-lymphocyte serum (ALS) did not appear to contain cytotoxic antibodies directed against membrane antigens on guinea pig lymph node lymphocytes. Furthermore, the anti-lymphokine serum did not inhibit the formation of spontaneous T rosettes nor significantly depress lymphocyte response to mitogens. The anti-lymphokine serum totally suppressed the delayed skin reactivity to PPD and contact sensitivity to DNCB when injected intradermally around the site of antigen challenge. By contrast, intradermally injected ALS did not appear to suppress the PPD response in sensitized guinea pigs. Intravenously and i.p. administered anti-lymphokine serum was somewhat less effective in suppressing the delayed skin response to PPD. The intradermal injection of the antiserum had no effect on nonspecific inflammation evoked by turpentine-olive oil or on the extravasation of circulating Evans blue evoked by intradermally injected histamine. Histologic examination of 24-hr DNCB-induced skin lesions from sensitized guinea pigs treated with intradermally injected anti-lymphokine serum showed marked reduction of mononuclear infiltration of the dermis and of epidermal lesions, as compared with skin sites taken from sensitized animals pretreated with normal goat serum. The anti-lymphokine serum injected i.v. also markedly reduced the perivascular infiltration of the dermis and subcutis in skin reaction sites from sensitized animals challenged with PPD. Intravenous treatment with ALS for 3 consecutive days caused extensive depletion of the paracortical areas of peripheral lymph nodes whereas treatment with normal serum and anti-lymphokine serum caused no such depletion. It is proposed that the anti-lymphokine serum is directed against activated lymphocyte products, one of them being MIF. These products are involved in the mediation of delayed hypersensitivity reactions. This is in marked contrast to ALS, the suppressive action of which appears to be central rather than peripheral.     

Effect of hyperimmune bovine colostrum raised against Cryptosporidium parvum on infection of guinea pigs by Cryptosporidium wrairi.

Oocysts shedding was markedly reduced in guinea pigs inoculated intraintestinally with Cryptosporidium wrairi sporozoites that had been incubated with hyperimmune bovine colostrum raised to C. parvum when compared with shedding in guinea pigs inoculated with sporozoites incubated in either non-immune bovine colostrum or buffered saline. However oocyst shedding was apparently not reduced in guinea pigs inoculated by gavage with oocysts of C. wrairi and subsequently treated twice daily per os with hyperimmune bovine colostrum. Similarly, oocyst shedding was apparently not reduced by oral treatment with hyperimmune bovine colostrum when treatment was begun simultaneously with inoculation of C. wrairi oocysts.     

The distribution and effects of intrapleural Corynebacterium parvum in mice

Summary After intrapleural (IPl) injection of ¹²⁵ I or fluorescein labelled C. parvum, most was confined to the pleural and mediastinal spaces. The pleural phagocytes and mediastinal lymph nodes were heavily labelled, but very little was found in the lung. The amounts of C. parvum taken up by the liver and spleen were less than after IV injection and splenomegaly was also less after IPl than IV injection. A large proportion (>90%) of cells in pleural washouts following IPl C. parvum was activated macrophages which inhibited, nonspecifically, the growth of tumor cells in vitro. No similar activity was detected after IV C. parvum. IPl injection of C. parvum mixed with irradiated tumor cells conferred strong, specific systemic immunity against tumor challenge, and this immunity was also demonstrable using mediastinal lymph node cells in a Winn assay. The immunity resulting from IV C. parvum and IPl irradiated tumor cells was significantly lower. IPl C. parvum has been compared with IV C. parvum for its effect against tumors growing either in the lung or pleural cavity. Tumors growing in the pleural cavity were inhibited more effectively by IPl than IV C. parvum. With tumors growing in the lung (caused by tumor cells injected IV), although IV C. parvum was more effective at reducing the number of lung nodules during the first two weeks, the mice consistently survived longer after IPl C. parvum.M.T.S. is a member of the Ludwig Lung Cancer Study Group. The present work arose out of discussions with other members of the group and is presented on their behalf. The study group is: M. Kaufmann, J. Stjernswärd (Ludwig Institut for Cancer Research, Lausanne Branch, Switzerland), M. Zelen, K. Stanley (Frontier Science and Technology Research Foundation, Inc. Amherst, New York, USA), D. S. Freestone, R. Bomford, M. T. Scott, T. Priestman (The Wellcome Research Laboratories, Beckenham, England), C. Mouritzen, G. Ahlbom (Dept. of Thoracic and Cardiovascular Surgery, Aarhus Kommunehospital, Aarhus, Denmark), N. Konietzko, D. Greschuchna (Ruhrland Klinik, Essen-Haidhausen, Germany), P. Hilgard (Innere Klinik und Poliklinik [Tumorforschung] Essen, Germany), J. Vogt-Moykopf, D. Zeidler, H. Toomes (Thoraxchirurgische Spezial-Klinik, Heidelberg-Rohrbach, Germany), F. Krause, R. Rios (Thoraxchirurgische Abt., Fachkrankenhaus für Lungen-und Bronchialerkrankungen, Löwenstein, Germany), J. Orel, M. Benedik, B. Hrabar (Clinical Center, Dept. of Thoracic Surgery, Ljubljana, Yugoslavia), S. Plesnicar (The Institute of Oncology, Ljubljana, Yugoslavia), H. A. Rostad, J. R. Vale (Rikshospital, Oslo, Norway), S. Hagen, S. Birkeland, (Ulleval Hospital, Oslo, Norway), T. Harbitz, R. Nissen-Meyer (Aker Hospital, Oslo, Norway), L. Rodriguez, V. O. Björk, K. Böök (Karolinska Sjukhuset, Thoracic Clinic, Stockholm, Sweden), E. Gradel, J. Hasse, P. Holbro (Kantonsspital, Thoraxchirurgische Klinik, Basel, Switzerland), L. Eckmann (Tiefenauspital, Chir. Univ.-Klinik, Bern, Switzerland), B. Nachbur, T. Liechti (Inselspital, Dept. of Thoracic and Cardiovascular Surgery, Bern, Switzerland), H. Cottier (Inst. of Pathology, Inselspital, Bern, Switzerland), W. Maurer, M. Kaufmann, P. Froelicher (Bürgerspital, Surgical Dept., Solothurn, Switzerland), H. Denck, N. Pridun (Krankenhaus der Stadt Wien-Lainz, Chir. Abt., Vienna, Austria), K. Karrer (Institute for Cancer Research, University of Vienna, Austria)Visiting Investigator. Recipient of an American Cancer Society Fellowship