Journal of Bacteriology & Parasitology

Cooperative Cytotoxins: A New Look at an Old Observation of Bacterial Crosstalk

Regina Linder^1* and Camille D. McIntyre^{2 1}Department of Medical Laboratory Sciences, Hunter College, Manhattan, New York, USA
²Department of Medical Laboratory Sciences, Hunter College, Manhattan, New York, USA
^*Correspondence: Regina Linder, Department of Medical Laboratory Sciences, Hunter College, Manhattan, New York, USA, Tel: 917 696 9460, Email:

Received: 16-Sep-2019 Published: 10-Oct-2019, DOI: 10.35248/2155-9597.19.10.362

Introduction

Beginning with the CAMP reaction of group B streptococci (GBS), and named for the investigators who described it in 1944 [1], many common bacterial pathogens and commensals were found to participate in “cooperative” cytotoxicity or “synergistic” hemolysis. The reaction is typically visualized as a zone of enhanced hemolysis between two plated species on blood agar (often sheep blood), but not seen around either organism separately. Alternatively, some bacterial pairs may cause inhibition of lysis of one species by the other [2]. Mammalian cells other than erythrocytes can be damaged [3]. The biochemical mechanisms underlying many of these relationships were characterized in earlier decades following the purification of the individual cytotoxins, and usually involve enzymatic hydrolysis of (or binding to) membrane lipid components of the target cells (phospholipids or sterols) with paired agents, acting sequentially [2,3].

Typically one of the agents (the phospholipase) is considered to hydrolyze a superficial membrane component, sphingomyelin or phosphatidylcholine (lecithin). Sphingomyelin predominates in the outer leaflet of sheep erythrocyte membranes [4], allowing access of the partner toxin to its molecular target residing deeper in the bilayer [5]. Table 1 illustrates some key examples, together with their known mechanisms of action.These commonly observed in vitro partnerships can be considered together with current concepts of the microbiome (the resident microbial communities on human tissues [6] thus highlighting cooperative cytotoxicity as a means of inter-species communication in hosts.

Table 1: Cooperative/Inhibitory cytotoxins of some common bacterial pathogens.

Tissue and wound infections often reflect mixed flora with varied relationships to oxygen, and include known cooperative (and inhibitory) partner species [7,8] similarly providing a plausible model for in vivo activity. Based on typical environments of participating species, and known mechanisms of toxin action, collaborative membrane damage among resident and pathogenic microorganisms can be inferred to be active in humans and animals, in health and disease.

Results demonstrate that cooperative and inhibitory hemolysis among species does not require that both strains grow concomitantly. Pre-grown strict anaerobes (clostridia) collaborate to lyse erythrocytes with strict aerobes grown the following day, after transfer to air incubation. Facultative species collaborate following sequential or concomitant growth, in either anaerobic or aerobic conditions. This plasticity reflects the ubiquity of many of these microorganisms as normal resident flora on tissues, as well as isolates from a wide range of infectious disease scenarios [9].

Materials and Methods

Bacterial strains; growth and maintenance

Bacterial strains used in this study derive either from long maintained culture collections of the authors [3], the Medical Laboratory Sciences Dept. of Hunter College, or clinical strains (designated KS) isolated from patient specimens at the VANY Harbor Healthcare System, by our late colleague Dr. Kuldip Sandhu, in 2007-8. Where strains were derived from public culture collections, i.e. ATCC, the Collection’s designation is retained. These include Rhodococcus equi 33701 and Corynebacterium pseudotuberculosis 19140. Strains are stored as suspensions in rabbit blood at -80°C. Routine growth of strains is in trypticase soy broth or agar (Difco), at 37°C in air, or Gas Pak (BBL) for Clostridium perfringens.

Demonstration of cooperative hemolysis or inhibition

Cross-streaking conditions have been described [2,3]. Cooperative cytotoxicity between wells inoculated with suspensions of known colony forming units (cfu) were carried out following inoculation as indicated and incubation in stated conditions. Documentation follows incubation, by digital photography.

Results

Enhanced as well as inhibitory hemolytic patterns of some pathogenic and commensal bacterial pairs, following sequential aerobic and anaerobic incubation are illustrated in Figure 1, a diameter streak of Clostridium perfringens (A) was pre-incubated overnight anaerobically, followed by perpendicular application of the test species [1-5] and continued incubation in air (24 hrs, 37°C). For B, Staphylococcus aureus is the diameter streak, and was plated concurrently with test strains 1-5, followed by 24 hrs at 37°C in air. The active toxin of C. perfringens is phospholipase C (α- toxin, lecithinase), and for S. aureus, sphingomyelinase C (β-toxin), and similar patterns of hemolysis are typically shown for both conditions and phospholipid sources [4]. The coryneform bacteria, C. glucuronolyticum and R. equi, together with S. agalactiae (GBS) showed enhanced hemolysis, in distinctive shapes. A. haemolyticum and C. pseudotuberculosis, were somewhat inhibitory to partial hemolysis by the diameter streak.

Figure 1. Cooperative hemolysis by human and animal pathogens grown in proximity to diameter plating on sheep blood agar of (A) C. perfringens (after prior anaerobic cultivation), and (B) S. aureus (plated concurrently in air): 1. Corynebacterium glucuronolyticum, 2. Rhodococcus equi, 3. Streptococcus agalactiae (GBS), 4. Arcanobacterium haemolyticum, 5. Corynebacterium pseudotuberculosis.

Hemolytic enhancement can also be viewed when standard bacterial inocula (2-3 × 104 cfu) were placed in agar wells from which diffusible proteins interact on erythrocytes. The resulting hemolytic zones can be compared in a semi-quantitative fashion. Figure 2 shows the aerobic collaboration between S. aureus (β- toxin+ in central wells, panel (A) and three strains of GBS isolated from clinical specimens (diabetic foot ulcers) with copathogen, Staphylococcus aureus. The size of hemolytic zones were qualitatively similar (ranking KS17>KS7>KS15) to those that developed when the same strains were incubated anaerobically opposite I (panel B). In neither case was S. aureus (included as a control, KS13) active in enhanced hemolysis.

Figure 2. Cooperative hemolysis between (A) Staphylococcus aureus (central wells) or (B) Clostridium perfringens (central wells) and duplicate peripheral wells containing three strains of GBS (KS7 top left, KS17 bottom left, KS15 bottom right) incubated in (A) Air and (B) Anaerobically. A clinical strain of S. aureus (KS13, top right) serves as a negative control.

Figure 3 shows cooperative hemolysis between C. perfringens (a strict anaerobe) and R. equi (a strict aerobe) when central wells inoculated with the clostridium (2-3 × 104 cfu, panel A) were preincubated anaerobically for 24 hrs, 37°C, followed by addition of 4 strains of R. equi, and incubation continued in air for an additional 24 hrs. In panel B, the same peripheral strains of R. equi are inoculated and incubated in air for 24 hrs 37°C, followed by inoculation of C. perfringens in central wells, and anaerobic incubation continued for 24 hrs at 37°C. Degree of hemolysis is similar for the two sequences, although the shape of hemolytic zones varies, the weakest cytotoxin of the R. equistrains (7502) did not produce visible hemolysis when anaerobic cultivation preceded air incubation.

Figure 3. Cooperative hemolysis between Clostridium perfringens (central wells) and 4 strains of Rhodococcus equi in duplicate peripheral wells (33701 top left, 1621 bottom left, C top right, 7502 bottom right), following inoculation of (A) Central wells anaerobically followed by peripheral wells in air; and (B) Peripheral wells inoculated and incubated in air prior to addition of C. perfringens (central) and anaerobic incubation.

Taken together, Figures 1-3 demonstrate the plasticity of cooperative cytotoxicity with regard to sequence of growth and oxygenation. Toxins (phospholipases, cholesterol oxidase, CAMP factor) released into the agar retain their activity and ability to collaborate, even when conditions are changed to prevent continued growth of the bacteria that produced them. Thus, PLC secreted from C. perfringens during anaerobic incubation retains its robust ability to lyse erythrocytes when exposed to the cholesterol oxidase of R. equi grown subsequently in air (Figure 3) as shown in early studies [4]. It should be noted that facultative bacteria (i.e. GBS clinical isolates and coryneforms) retained the ability to cooperate even when all strains were cultivated anaerobically, and when growth was sparse (Data not shown).

Discussion

From its first description in 1944, the cooperative cytotoxic relationship between S. aureus and GBS (the CAMP reaction) proved a valuable diagnostic test to quickly categorize hemolytic streptococci important in both human and animal health [1,2]. Similar observations followed among other pathogens and commensals, providing assistance with presumptive lab identifications. However, little attention has been directed to the function or evolutionary benefit to the organisms that participate. Considering the large number of such relationships, as well as the property of these reactions to occur across the barrier of anaerobiosis, it is proposed that cooperative cytotoxicity be considered an example of bacterial “crosstalk”, the complex array of microbial, immunological and environmental factors that impact host outcomes [9,10].

In considering the identity of organisms that participate in cooperative relationships themes emerge: Coryneform bacteria (the genus Corynebacterium, and related Gram-positive pleiomorphic rods), are associated with ruminant and other animal infections, as well as opportunistic disease in humans. An example is R. equi, a frank pathogen of foals, is able to infect immunocompromised humans exposed to equine environments [11].

Infections typically occur on mucosal surfaces with a mixed resident flora, including S. aureus and C. perfringens. Listeria monocytogenes is isolated from a wide variety of animals, birds and fish. Infection in humans is typically foodborne, and potentially serious in high risk hosts. Acinetobacter is a diverse genus increasingly recognized as a hospital opportunist, and resistant to multiple antibiotics. In general, pathogenesis among these species is opportunistic, and may derive from the commensal flora or from exogenous exposure to animals or soil. Infection most often develops on contaminated, rather than sterile, sites [12].

Cooperative cytotoxicity is not limited to bacteria. The observation that common dermatophyte fungi display synergistic hemolysis in partnership with staphylococci and Listeria [13] expands the potential for co-toxicity. The PLD of C. pseudotuberculosis is identical with the toxin of the brown recluse spider [14] introducing the opportunity for communication with sources other than microorganisms. Cytotoxins with activity profiles similar to those of bacteria have been identified from a range of biological sources, including snakes, insects, and sea anemones [15].

Microbial pathogenesis involving strictly anaerobic bacteria, including Clostridium perfringens often depends upon collateral tissue damage or co-infection. When colonized tissue (e.g. intestinal, oral, and skin) is damaged by mechanical force or infection with aerobic or facultative bacteria, the reduced oxygen tension that results supports growth of resident anaerobes, out of proportion to their usually balanced numbers [12].

Polymicrobial isolations are commonplace, and include bacteria with varied relationships to oxygen [9]. Molecular analysis of rRNA and similar markers from infected and uninfected chronic wounds reveals highly diverse flora, with dozens of species, facultative and anaerobic, and often including S. aureus [10]. The potential for cooperative interactions is extensive, albeit refractory to direct demonstration.

We suggest that cooperative cytotoxicity be considered among mechanisms of inter-species communication, providing enhanced ability to damage host tissue, increase inflammation, and introduce specific metabolic products associated with harm to humans. For example, ceramide (exposed on target cell membranes when sphingomyelin is hydrolyzed) is associated with an increasing range of physiological harm, including to mitochondrial function [16], and is recognized in the pathogenesis of Parkinsonism [17]. Oxidized sterols, a product of the sequential action of phospholipase and cholesterol oxidase have long been implicated in complex cardiovascular pathology [18]. In an earlier experimental protocol, cholesterol oxidase was directly lethal to rabbits rendered atherosclerotic by diet [19].

It should also be noted, in the literature and in our hands, weak cooperative hemolysis is a wider phenomenon than the well characterized robust cytotoxic pairings here discussed. Many streptococci, including group A and the pneumococcus are examples [3,4]. While not deemed useful in the identification of these species, cytotoxicity should not be ruled out among the numerous virulence determinants of these pathogens. Further, it is significant to recall that red blood cells in these studies are a convenient model for plasma membranes of tissues and organs that may be exposed. The combined action of paired toxins was shown to act on cultured mammalian cells [5].

Cytotoxic partnerships among microorganisms are examples of the recently articulated concept, the “pathobiome” [20]. According to Bass et al., it is the complex sum of interactions among commensal, disease producing and host elements that influence the incidence and outcome of infectious diseases in humans, plants and animals [20]. Consistent with this approach, the microbiome, the mixture of microbial residents deemed critical for maintaining good health [21] has been shown extend to infected tissue. In a recent study of diabetic wounds, Kalan et al., using molecular analysis, demonstrated the microbial diversity of persistent lesions, including aerobes and anaerobes, corynebacteria and very commonly S. aureus [9].

Table 1 illustrates a range of bacterial species that participate in robust cooperative and inhibitory cytotoxic partnerships, together with the enzymatic or binding products that have been shown to mediate the reactions. In order to highlight the significance of cytotoxic cooperation, the typical pathology of species is indicated [22,23].

Conclusion

Thus, the numerous examples of cooperative (and inhibitory) cytotoxicity described decades ago highlight a subset of microbial and mammalian interactions that reflect contemporary concepts of health and disease.

REFERENCES

1. Christie R, Atkins NE, Munch-Petersen E. A note on a lytic phenomenon shown by group B streptococci. Aust J Exp Biol Med Sci. 1944;22(3):197-200.
2. McFaddin JF. Biochemical Tests for Identification of Medical Bacteria. (3rd edn) Williams Wilkins publishers, Philadelphia. 2000. pp. 35-56.
3. Linder R. Alteration of mammalian membranes by the cooperative and antagonistic actions of bacterial proteins. Biochim Biophys Acta. 1984;779(4):423-435.
4. Gubash SM. Synergistic hemolysis phenomenon shown by an alpha-toxin-producing Clostridium perfringens and streptococcal CAMP factor in presumptive streptococcal grouping. J Clin Microbiol. 1978;8(5):480-488.
5. Linder R, Bernheimer AW. Action of bacterial cytotoxins on normal mammalian cells and cells with altered membrane lipid composition. Toxicon. 1984;22(4):641-651
6. Sultan A, Kong EF, Rizk AM, Jabra-Rizk MA. The oral microbiome: A lesson in coexistence. PLoS Pathog. 2018;14(1):e1006719.
7. Linder R, Bernheimer AW. Enzymatic oxidation of membrane cholesterol in relation to lysis of sheep erythrocytes by corynebacterial enzymes. Arch Biochem Biophys. 1982;213(2):395-404.
8. Percival SL, Malone M, Mayer D, Salisbury AM, Schultz G. Role of anaerobes in polymicrobial communities and biofilms complicating diabetic foot ulcers. Int Wound J. 2018;15(5):776-782.
9. Kalan LR, Brennan MB. The role of the microbiome in nonhealing diabetic wounds. Ann N Y Acad Sci. 2019; 1435(1):79-92.
10. Bowler PG, Davies BJ. The microbiology of infected and uninfected leg ulcers. Internal J Dermatol. 1999; 38(8):573-578.
11. Linder R. Rhodococcus equi and Arcanobacterium haemolyticum: Two “Coryneform” bacteria increasingly recognized as agents of human infection. Emerg Infect Dis. 1997;3(2):145-153.
12. Tille P. Bailey & Scott Diagnostic Microbiology. (14 edn) Elsevier, St. Louis, Missouri. 2017. pp: 552-580.
13. Schaufuss P, Brasch J, Steller U. Dermatophytes can trigger cooperative (CAMP-like) hemolytic reactions. Br J Dermatol. 2005;153(3):584-590.
14. Bernheimer AW, Campbell BJ, Forrester LJ. Comparative toxinology of Loxosceles reclusa and Corynebacterium pseudotuberculosis. Science. 1985;228(4699):590-591.
15. Linder R, Bernheimer AW. Comparative toxinology of bacterial and invertebrate cytolysins. In: Alouf JE, Freer JH (eds). Sourcebook of Bacterial Protein Toxins. Academic Press, London, 1991: pp: 101-115.
16. Novgorodov SA, Voltin J, Gooz MA, Li L, Lemasters JJ, Gudz TI. Acid sphingomyelinase promotes mitochondrial dysfunction due to glutamate-induced regulated necrosis. J Lipid Res. 2018;59(2):312-329.
17. Plotegher N, Bubacco L, Greggio E, Civiero L. Ceramides in Parkinson’s Disease: From recent evidence to new hypotheses. Front Neurosci. 2019;13:330.
18. Tang HY, Wang CH, Ho HY, Wu PT, Hung CL, Yeh YH, et al. Lipidomics reveals accumulation of the oxidized cholesterol in erythrocytes of heart failure patients. Redox Biol. 2018;14:499-508.
19. Bernheimer AW, Robinson WG, Linder R, Mullins D, Yip YK, Cooper NS, et al. Toxicity of enzymically- oxidized low-density lipoprotein. Biochem Biophys Res Commun. 1987;148(1):260-266.
20. Bass D, Stentiford GD, Wang HC, Koskella B, Tyler CR. The pathobiome in animal and plant diseases. Trends Ecol Evol. 2019;34(9):7-12.
21. Ebersole R. You are what you eat. Food and Environment Reporting Network. 2018.
22. Bernheimer AW, Linder R, Avigad LS. Nature and mechanism of action of the CAMP protein of Group B streptococci. Infect Immun. 1979;23(3):838-844.
23. Barksdale L, Linder R, Sulea IT, Pollice M. Phospholipase D activity of Corynebacterium pseudotuberculosis (Corynebacterium ovis) and Corynebacterium ulcerans, a distinctive marker within the genus Corynebacterium. J Clin Microbiol. 1981;13(2):335-343.