Abstract
Background: Bacteriophage therapy is considered one of the most attractive alternatives to antibiotic treatment, which may be significant due to the rising number of antibiotic-resistant bacterial strains. Patients with cancer frequently suffer bacterial infections resulting from immunosuppression caused by anticancer treatment; thus they constitute a considerable group of patients subjected to phage therapy. In this study, we investigated the influence of bacteriophages on the migration of human leukemia (HL-60) cells. Results of these studies provide data regarding phage treatment of patients with cancer, especially with this type of leukemia. Materials and Methods: The influence of phage preparation on migration of HL-60 leukemia cells was evaluated with BD Bioscience Migration Chambers. Results: Bacteriophages have no influence on migration of HL-60 cells. The only phage preparation which stimulated migration of HL-60 cells was Staph.liz, specific to S. aureus, however, the molecular basis of these interactions cannot be currently explained. Conclusion: Results of our studies may be in line with previous data indicating that phage therapy is safe for patients with cancer.
Bacteriophages are viruses that infect bacteria. Attempts to use bacterial properties of phages have been made almost since their discovery at the beginning of the 20th century. Treatment with phage preparations, known as phage therapy, was however marginalized after the discovery of penicillin and the rapid development of the antibiotic industry. Today, while the number of antibiotic-resistant bacterial strains is rapidly increasing, phage therapy is considered one of the most promising alternatives to classical treatment (1-4).
The efficacy of phage therapy has been demonstrated in the treatment and prevention of bacterial infections caused by numerous bacterial species (5-12), including multidrug-resistant species (13-17).
Bacteriophage therapy has several advantages over antibiotic treatment. Bacteriophages are extremely specific viruses (they infect only the host bacterial strain or a narrow range of related bacteria); thus they affect only target bacteria, not affecting the natural microflora. Bacteriophages are ‘auto-dosing’, which means they multiply specifically where their host is located (requiring a relatively high bacterial density) and establish themselves the phage dose (18, 19). Bacteriophages are able to degrade at least some biofilms (which are more often resistant to antibiotics than planktonic bacteria) (20). Phage preparations may be versatile in different application forms and routes of administration (3, 21, 22). Bacteriophage preparations are obtained via a cheap and non-hazardous process. Phages may be isolated from readily available sources such as sewage and water (18).
There are some limitations of phage therapy. The high specificity of phages implies the necessity for precise selection of active viruses. The problem of this high specificity may however be circumvented by application of phage cocktails, which are combinations of a few phages. This results in extension of the activity spectrum (23, 24). Phage resistance is also possible (25). Interestingly, mutations causing phage resistance sometimes simultaneously cause impairment of bacteria vitality, which may result in loss of pathogenicity (18). Phage preparations are based on crude bacterial lysates; thus they contain bacterial residues (including endotoxins) which may have negative effects on the body during therapy. One problem may be delivery of antibodies against bacteriophages during therapy (26-28). It is worth emphasizing that phage therapy is an ‘experimental’ form of treatment, so despite promising results and no proven side-effects, extensive studies of all aspects of phage treatment need to be conducted.
Bacteriophages are believed to have no natural tropism to eukaryotic cells; however, it was found that phages may interact with mammalian cells and those interactions may be of great importance (29-40). The first report regarding interactions between bacteriophages and cancer cells comes from 1940, when Bloch observed the ability of bacteriophages to accumulate in Ehrlich carcinoma tissue. Moreover, phages seemed to inhibit tumor growth (41). Since then, interactions between bacteriophages and cancer cells have repeatedly been reported. Kantoch and Mordarski described in vitro binding of T2 phages to Ehrlich carcinoma and amytal sarcoma (42). Further studies indicated phage influence on migration of murine melanoma in vitro (43, 44) and both melanoma and Lewis lung cancer in vivo (45).
Data regarding phage therapy in patients with cancer date back to the 1980s (46, 47). Presently, cancer patients constitute a significant group of patients who undergo phage therapy, as bacterial infections frequently result from immunosuppression caused by anticancer chemotherapy. An open question remains whether bacteriophage preparations may be safely applied to such patients.
In this study, we evaluated the in vitro influence of phage preparations on the migration of human leukemia HL-60 cells. Compared to previous studies, we have greatly broadened the range of phage preparations used in the experiments (24 phage preparations from three groups of phages, including crude phage lysates and purified phage preparations). We also used a new group of phage preparations (the majority were preparations used in phage therapy, unlike T-phages most commonly used in previous studies but extremely rarely applied in therapy). Increased mortality of patients with leukemia is not related to intense migration of leukemia cells, but rather to their escalated amplification. In spite of that, we decided to investigate phage influence on migration of leukemia cells, as this process has been most often studied in other types of cancer. We also chose to compare results with those of our previous studies, in which we investigated the influence of phage preparations on the migration of human phagocytes. HL-60 is a promyelocytic leukemia cell line; thus an interesting issue was whether the influence of phage on migration of this cancer cell line is analogous to that for other cancer types or granulocytes (whose migration was not influenced by phage preparations) (Kurzepa-Skaradzinska et al., data not shown).
Materials and Methods
Bacteriophages and bacteria. Bacteriophages and bacteria were obtained from the Polish Collection of Microorganisms [Institute of Immunology and Experimental Therapy (IIET), Polish Academy of Sciences (PAS)]. In the experiments we used bacteriophage preparations most frequently applied in phage therapy at the Phage Therapy Centre (IIET) in Wroclaw. Phages were propagated with bacteria from the IIET Bacterial Collection. Bacteriophages with their host bacteria are listed in Table I.
Bacteriophage lysates were prepared according to the procedure described by Carlson and Miller (48). Purified preparations of phages T2, T4, HAP1, A5 and A3/R were obtained according to the method of Boratynski et al. (49). Endotoxin levels in preparations were determined with the Limulus Amebocyte Lysate Test Kit (LAL) (BioWhittaker, Lancaster, Massachusetts, USA). The phage plaque in all preparations was determined with the double agar method (50). In most experiments, we used preparations with a phage plaque of 1×109 pfu/ml. To determine whether phage preparations with lower phage plaque influence migration of human phagocytes, we also used lysates and purified phage preparations with phage plaque 1×107 pfu/ml and 1×105 pfu/ml.
Lipopolysaccharide. Lipopolysaccharide (LPS) of Escherichia coli B was kindly provided by Dr. Kinga Switala-Jelen according to the procedure described by Westphal (51). Purified phage preparations contain a certain amount of bacterial residues; thus LPS solution was used as a control of the potential activity of this contamination in purefied phage preparations.
Tumor cells. The HL-60 cell line was obtained from the Collection of Reference Cell Lines from the Laboratory of Anticancer Experimental Therapy (IIET, PAS). Cells were maintained in in vitro culture in RPMI medium with 10% fetal bovine serum (FBS), 4 mM L-Glu, glucose and sodium pyruvate. They were cultivated in an incubator at 37°C (in a humid atmosphere (with 5% CO2).
Preparation of bacteriophage lysates and bacterial control lysates. One milliliter of 20-h bacterial culture was introduced into two flasks with 50 ml of culture medium (CM), then 1 ml of phage preparation (lysates from Table I) was added to one of the flasks. At 30 min intervals, the OD 600 of the culture with phage present was measured. When the OD 600 began to drop (bacterial lysis occurred), the second culture was disrupted by ultrasound, 100 μl of lysozyme was added, then the culture was placed for 30 min on ice followed by at least 24 hours at a temperature of −80°C. The culture with phage addition was incubated until the OD 600 of the culture was close to zero. Both cultures were filtered with 0.22 μm antibacterial filters and stored at 4°C. The culture with phage addition was phage lysate and the second culture was bacterial control lysate (BCL).
Influence of phage preparation on migration of HL-60 cells. The influence of phage preparation on migration of HL-60 leukemia cells was evaluated with BD Bioscience Migration Chambers (BD Biosciences, San Jose, California, USA).
Preparation of migration chambers: The outer membrane of the insert was covered with 20 μl of human fibronectin solution (100 μg/ml) and dried at 37°C for 30 min. The inner membrane was then covered with an equivalent volume of fibronectin and dried under the same conditions. The inserts and the wells were rinsed twice with mQ water, refilled with 500 μl of albumin solution and kept at 37°C for 15 min. Rinsing with mQ water was then repeated once.
Migration of HL-60 cells: HL-60 cells were incubated with preparations [phage preparation, (PBS), (CM), LPS or BCL] for 1 h (37°C, 5% CO2). Simultaneously, equivalent volumes of CM were incubated with the same preparations under the same conditions. Five hundred microliters of CM with preparations were introduced into the wells of a 24-well culture plate and 500 μl of cell suspension were introduced into the inserts. Inserts were put into the corresponding wells (with the same preparation) and the plate was maintained for 1 h (37°C, 5% CO2). The final densities of cells and phage preparations in the upper migration compartment were 1×106 cells/ml and 1×108 pfu/ml respectively. The phage titer in the well was equivalent to that in the insert suspension (no chemotactic interactions). The influence of phages with lower phage titer was also studied. The final phage densities in solutions inside inserts and those in wells were 1×106 pfu/ml and 1×104 pfu/ml. The cell density was the same as in the other experiments (1×106 cells/ml). PBS and LPS (with the same level as in purified phage preparation) were used as controls in studies of influence of purified phage preparations on migration of HL-60. PBS, CM and BCL were used as controls in studies of influence of phage lysates on migration of HL-60
Staining and counting of HL-60 cells: The cell suspension was removed from the inserts and the inner membrane was cleaned with cotton swabs in order to remove remaining cells. HL-60 cells which had migrated through the membrane were stained with Diff-Quick Set (Medion Diagnostics GmbH, Düdingen, Switzerland) (staining procedure according to producer's instructions). Cells were then counted in an optical microscope (magnification: ×400).
Statistical analysis. Data were analyzed with parametric or non-parametric tests, depending on the distribution of variables. Accordance with the normal distribution was evaluated with the Shapiro–Wilk test. If the distribution of variables was in accordance with the normal distribution, homogeneity of variance in all groups was evaluated with the Bartlett test; otherwise, the Fligner–Killeen test was applied. If variances were homogeneous, and in accordance with or slightly deviating from the normal distribution, data were analyzed with analysis of variance (ANOVA), and then with the Tukey multiple comparison test. Otherwise, data were analyzed with the Friedman test and then with the Wilcoxon test with Holm adjustment. If ANOVA indicated statistical significance, and statistically weaker multiple comparison tests did not produce unequivocal results (p>0.05 for each comparison), according to rules of statistical analysis, the largest difference between arithmetic means for which the ANOVA p-value was given was considered statistically significant (52).
Results
Influence of phages specific to E. coli on the migration of HL-60 cells. Phage lysates applied in phage therapy: Phage preparations did not influence the migration of HL-60 cells. Compared to PBS, CM and BCL, only the phage preparation OX1/DSM showed a tendency to inhibit migration of HL-60 cells; however, the results were not statistically significant.
Lysates of T-phages: Lysates of T-phages did not influence migration of HL-60 cells.
Purified T-phage preparations: Compared to PBS, purified T-phage preparations did not influence migration of HL-60 cells. Compared to LPS (which had a tendency to stimulate migration compared to PBS), phage preparations had a tendency to inhibit migration of HL-60 cells. However, the results were not statistically significant.
Influence of phages specific to Staphylococcus sp. on migration of HL-60 cells. Phage lysates applied in phage therapy: Most phage preparations did not influence on the migration of HL-60 cells. The exception was the preparation of bacteriophages Staph.Liz/80, which, compared to BCL, stimulated migration of HL-60 cells (p=0.045; Tukey test, ANOVA) (Figure 1). Preparations of phages 676/Z and A3/R had a similar tendency, but the differences were not statistically significant. Preparations of phages A5/80 and P4/6409 had a tendency to inhibit migration of HL-60 cells, but the differences were not statistically significant.
Purified A5/80 phage preparation: A5/80 purified phage preparation influenced migration of HL-60 cells to an extent comparable to that of PBS. Simultaneously, like purified T- phages, A5/80 phage preparation had a tendency to inhibit migration of leukemia cells compared to the LPS group. However, the differences were not statistically significant.
Influence of phages specific to Pseudomonas aeruginosa on the migration of HL-60 cells. Phage lysates applied in phage therapy: Most phage preparations of this group did not influence migration of HL-60 cells. Phage lysate 1214 inhibited migration compared to PBS and CM; however, the effect was comparable to that of BCL. This indicates activity of bacterial residues in the preparation. A similar effect was induced by the preparation of bacteriophage 119x: it inhibited migration of HL-60 cells compared to CM (p=0.01; ANOVA, Tukey test); compared to PBS, only a tendency to inhibit migration was observed (p=0.062; ANOVA, Tukey test). There were no significant differences between the effect induced by the phage preparation and its BCL, which again indicates activity of bacterial residues in the preparation.
Influence of phage preparations with different phage titers on migration of HL-60 cells. Lysate of T4 phage: We evaluated the influence of T4 phage lysates with different phage titers (1×105 pfu/ml; 1×107 pfu/ml and 1×109 pfu/ml) on migration of HL-60 cells. None of the preparations influenced migration of HL-60 cells.
Purified T4 phage preparation: We evaluated the influence of purified T4 phage preparation with different phage titers (1×105 pfu/ml; 1×107 pfu/ml and 1×109 pfu/ml) on migration of HL-60 cells. None of the preparations influenced migration of HL-60 cells.
Discussion
The results presented here concern the influence of bacteriophages on the migration of human leukemia cells (HL-60). Reports regarding interactions between bacteriophages and this cell line are extremely scarce. Furthermore, most publications regarding interactions between bacteriophages and cancer cells include studies of only a few phage preparations. In our study, we significantly broadened the pool of preparations used in the experiments; moreover, the majority consisted of phage preparations directly applied in phage therapy, which we believe has practical consequences. We selected three groups of phages: specific to E. coli, S. aureus and P. aeruginosa, as infections caused by these bacteria are relatively frequent. For each group, we selected six phage preparations most commonly administered to patients at the Phage Therapy Unit in Wroclaw. We also used three T-phages (T2, T4, HAP1) which are rarely applied in phage therapy but are often used in studies regarding interactions between bacteriophages and mammalian cells (31-35, 38-40, 43, 45, 53, 54).
Most phage preparations used in our study did not influence migration of HL-60 cells. The only phage preparation which stimulated migration of leukemia cells was Staph.liz/80, specific to S. aureus. The molecular basis of interactions between bacteriophages and HL-60 cells cannot be currently explained due to extremely scant and contradictory data regarding the molecular structure and biology of these viruses. We may assume that the role of phage proteins related to phage specificity is limited in the described interactions as the A5 bacteriophage, which has the same host bacteria as Staph.liz, does not influence the migration of HL-60 cells.
The hypothesis regarding interactions between bacterio-phages and mammalian cells assumes binding of KGD tri-peptide (a fragment of capsid proteins of some phages) with integrins in the membranes of certain mammalian cells (34). This hypothesis may be confirmed by the fact that murine melanoma migration is inhibited by T-phages: T4 and HAP1 (containing the KGD motif in the main capsid protein gp24). T4 mutant HAP1, deprived of the Hoc protein, protruding from the phage head and constituting a steric barrier for interactions with gp24, more strongly inhibits migration of B16 (43, 45, 54). In our study, neither T4 phage nor HAP1 influenced migration of HL-60 cells. Likewise, T2 phage, with a similar structure to T4, also deprived of the Hoc protein, did not influence on the migration of HL-60 cells. This could mean that the role of gp24 is limited in the described interactions. The results presented here show that phages may not interact with each type of cancer cell (phages T4 and HAP1 which inhibited migration of murine B16 melanoma and LLC (43, 45, 54) did not influence migration of HL-60 cells). The influence of bacteriophages on the migration of cancer cells may thus depend on the cancer cell line. The adhesivity of cancer cells may possibly be of some importance in phage–cancer cell interactions. Studies on the interactions between phages and different types of cancer cells (including other non-adhesive cancer cell types) need to be carried out as the interactions that occur undoubtedly constitute an interesting research field.
To investigate whether bacterial residues in phage preparations influence migration of HL-60 cells, we compared the activity of some phage lysates with corresponding purified preparations of phages T2, T4, HAP1 (specific to E. coli) and A5 (specific to S. aureus). Currently, it is not possible to obtain a purified preparation of phage specific to P. aeruginosa with high phage titer and good stability. The results of our study indicated that bacterial residues do not influence on migration of HL-60 cells. To confirm these data we investigated the influence of LPS isolated from E. coli and sonicates of gram-positive (S. aureus) and gram-negative (E. coli) bacteria on migration of HL-60 cells. Only LPS had a tendency to stimulate migration of cells; however, the data were not statistically significant.
It is known that some drugs may be effective at lower concentrations; thus, we investigated the influence of phage preparations with different phage titers on migration of HL-60 cells. We tested T4 lysate and T4 purified preparations with phage titers 1×105 pfu/ml, 1×107 pfu/ml and 1×109 pfu/ml. None of the phage preparation influenced migration of HL-60 cells.
Although there are no data indicating that increased mortality of patients with leukemia is associated with increased migration of cancer cells, we chose leukemia as a model line in our study to assess whether the influence of phages on HL-60 migration would be comparable to that observed in the case of other cancer types or there would be no interactions as between bacteriophages and phagocytes (Kurzepa-Skaradzinska et al., data not shown). The results of our study indicate that phages do not appear to influence migration of HL-60 cells (the effect is similar to that for phagocytes).
There are no data indicating any side effects of phage therapy. The results of our study support previous data indicating that phage therapy may be safe for patients with cancer. However, further studies are certainly required.
- Received January 8, 2013.
- Revision received February 18, 2013.
- Accepted February 19, 2013.
- Copyright© 2013 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved