Characterization of the hemolytic activity of Riemerella anatipestifer

Abstract

Riemerella anatipestifer infection causes serious economic losses in the duck industry worldwide. Acute septicemia and high blood bacterial loading in R. anatipestifer infected ducks indicate that R. anatipestifer may be able to obtain iron and other nutrients by lysing duck erythrocytes to support its rapid growth and proliferation in the blood. However, so far, little is known about the hemolytic activity of R. anatipestifer to duck erythrocytes. In this study, 29 of 52 R . anatipestifer strains showed hemolytic activity on duck blood agar, whereas all the tested dba⁺ (with hemolytic activity on duck blood agar) and dba⁻ strains created pores in the duck red blood cells, with 4.35–9.03% hemolytic activity in a liquid hemolysis assay after incubation for 24 h. The concentrated culture supernatants of all the tested R. anatipestifer strains and the extracted outer membrane proteins (OMPs) from dba⁺ R. anatipestifer strains showed hemolytic activity on duck blood agar. These results, together with the median lethal dose (LD50) of some dba⁺ and dba^- R. anatipestifer strains in ducklings, suggested that there was no direct relationship between the hemolytic capacity of R. anatipestifer on duck blood agar and its virulence.

Riemerella anatipestifer , the type species of the genus Riemerella in the family Flavobacteriaceae , mainly infects domestic ducks, geese, turkeys, and other birds. R. anatipestifer infection primarily caused an acute septicemic disease in younger birds and more chronic and localized lesions in older birds. It is a major disease challenging the duck industry worldwide, and accounts for significant economic losses [1]. However, the mechanisms underlying the pathogenesis of R. anatipestifer infection are not clearly understood.

The ability to acquire elemental iron is of pivotal importance to bacterial survival, proliferation, and the establishment of infection within their hosts [2, 3]. R. anatipestifer could not grow on LB agar, but could on LB agar with animal red blood cells [1, 4]. It suggested that R. anatipestifer could acquire iron and/or other nutrients from lysed erythrocytes for its growth and proliferation. In vivo, the acute septicemia [5] and high blood bacterial loading [6] in R. anatipestifer infected ducks, indicated that R. anatipestifer may be able to obtain iron and other nutrients by lysing duck erythrocytes to support its rapid growth and proliferation in the blood, and then bacteria could spread with bloodstream to the whole body rapidly, thus facilitating its systemic infection process in ducklings. However, few studies on the hemolytic activity of R. anatipestifer has been reported. Some R. anatipestifer strains showed hemolysis on 5 % bovine blood agar during 14 days examination [7], or on Columbia agar base with 7 % defibrinated sheep blood [8]. So far, the hemolytic activity of R. anatipestifer has not yet been investigated using the erythrocytes from its natural infection host – duck, and no hemolysin has been identified. We also did not know whether the hemolysis of R. anatipestifer was contact-dependent, or contact-independent through secreted hemolysins.

In addition, hemolytic activity has been linked to virulence in several bacterial species, such as Vibrio species [9], Mycobacterium tuberculosis [10, 11], and Bordetella pertussis [12]. Our previous studies showed that some iron utilization related proteins, such as siderophore-interacting protein [13], TonB-dependent receptor TbdR1 [14], TonB1 and TonB2 [15], were involved in the virulence of R. anatipestifer , and that R. anatipestifer can use hemoglobin and hemin as sole iron sources [13]. Since iron acquisition has been referred to as the ‘critical determinant,’ which decides the outcome of host-pathogen interactions [3], the hemolytic capacity of R. anatipestifer may be related to its virulence. However, the hemolytic activity of F. psychrophilum, which also belongs to the family Flavobacteriaceae, did not seem to be linked to its virulence [16]. So, whether the hemolytic capacity of R. anatipestifer directly corresponds to its virulence remains unclear. Therefore, in this study, we analysed the hemolytic capacity of different R. anatipestifer strains using the erythrocytes from its natural infection host ducks, and the relationship between the intensity of hemolytic activity and bacterial pathogenicity.

A total of 52 R . anatipestifer strains of different serotypes (Table S1, available in the online version of this article) were tested for their hemolytic activity on duck blood agar plates (LB agar base with 3.5 % duck blood, pH 7.5). The plates were then incubated at 37 °C under 5 % CO₂ for 24 h, followed by incubation at 4 °C overnight. The results showed that twenty-nine of the 52 (55.77 %) R . anatipestifer strains generated clear hemolytic zones (dba⁺), and 23 of the 52 strains (44.23 %) showed no hemolytic activity on duck blood agar (dba⁻). Among the dba⁺ strains, the hemolytic activities of strains 2, 3, 34, and 38 (Table S1) were stronger than those of the other strains. No direct relationship between the serotypes of the strains and their hemolytic activities was observed.

In addition, the hemolytic activity of five dba⁺ and five dba⁻ R. anatipestifer strains was further tested with a liquid hemolysis assay as described by Suzuki N et al. [2] with minor modifications. Briefly, R. anatipestifer bacteria were grown in TSB to mid-logarithmic phase, washed with 10 mM PBS (pH 7.2), and resuspended in Williams’ Medium E (Sigma-Aldrich, MO, USA) at a density of 1×10⁹ c.f.u. ml⁻¹. Equal volumes of the bacterial suspension and 2 % (v/v, in Williams’ Medium E) duck red blood cells were mixed in triplicate in a 96-well round-bottomed cell culture plate (Corning, NY, USA) and incubated at 37 °C for 24 h. The amount of hemoglobin released from the lysed erythrocytes in the culture supernatant was determined with spectrophotometry as the optical density at a wavelength of 540 nm (OD₅₄₀). The spontaneous lysis (negative control) was determined in the supernatant of duck red blood cells in Williams’ Medium E without bacteria under the same conditions, and in the positive control, the bacterial suspension in the assay was replaced with same volume of 0.1 % SDS (final concentration). The percentage hemolysis was calculated as: % hemolysis = (OD₅₄₀ of sample - OD₅₄₀ of negative control) / (OD₅₄₀ of positive control - OD₅₄₀ of negative control)×100. The results showed that all the tested R. anatipestifer strains, including dba⁻ strains 6, 17, 18, and 26, had hemolytic activity, causing between 4.3 and 9.0% lysis. The dba⁺ strains 2 and 3, which had the highest hemolytic activity on duck blood agar among the tested strains, displayed greater percentage hemolysis in the liquid assay than other strains, at 8.00 and 9.03 % respectively (Table S1).

To determine why the dba⁻ strains displayed hemolytic activity in the liquid hemolytic assay, the pelleted nonlysed duck blood cells were washed twice with Williams’ Medium E, and the morphology of the cells was observed under a Nikon Eclipse Ci phase contrast microscope (Nikon Corporation, Japan) to investigate whether both the dba⁺ and dba⁻ R. anatipestifer strains generated pores in the duck erythrocyte membrane. Duck blood cells alone in Williams’ Medium E were used as the negative control. The results showed that all the R. anatipestifer strains tested formed pores in the duck red blood cells at 12 h postincubation. However, the dba⁺ strains generated more pores in the duck erythrocyte membrane than the dba⁻ strains (Fig. 1).

Fig. 1.

All the R. anatipestifer strains tested generated pores in duck red blood cells, observed with phase contrast microscopy. Four dba⁺ strains 2, 3, 33 and 38, and five dba^- strains 6, 17, 18, 26 and 46 were respectively incubated with duck red blood cells at 37 °C for 12 h. The morphology of the duck red blood cells was observed under a Nikon Eclipse Ci phase contrast microscope. The observation through phase contrast microscopy showed that all the tested dba⁺ and dba^- R. anatipestifer strains generated pores in duck red blood cells, and the dba⁺ strains generated more pores in the duck erythrocyte membrane than the dba⁻ strains.

To determine whether the culture supernatants of hemolytic R. anatipestifer strains contained secreted hemolytic activities, the culture filtrates and 40-fold-concentrated culture filtrates of dba⁺ R. anatipestifer strains 1, 2, 3, 5, and 38, and dba⁻ strains 6, 17, 18, and 26, were prepared. The filtered culture supernatants from all the tested R. anatipestifer strains at mid-logarithmic phase showed no hemolytic activity on the duck blood agar, whereas the 40-fold-concentrated filtrate did show hemolytic activity. Interestingly, the culture supernatants of all the tested dba⁺ and dba⁻ R. anatipestifer strains generated many pores in the duck erythrocyte membrane, observed with phase contrast microscopy (Fig. 2). This suggested that the culture supernatants of dba⁺ and dba⁻ R. anatipestifer strains contained hemolysin.

Fig. 2.

The culture supernatants of all the tested dba⁺ and dba⁻ R. anatipestifer strains generated many pores in the duck erythrocyte membrane, observed with phase contrast microscopy. The culture supernatants from five dba⁺ R. anatipestifer strains 1, 2, 3, 5 and 38, and four dba⁻ strains 6, 17, 18 and 26, were incubated with duck red blood cells for 12 h. The observation through phase contrast microscopy showed that the culture supernatants of all the tested dba⁺ and dba⁻ R. anatipestifer strains generated many pores in the duck erythrocyte membrane.

To investigate whether the OMPs of R. anatipestifer are involved in hemolysis, the OMPs were extracted from mid-log-phase R. anatipestifer strains, including four dba⁺ strains 2, 3, 33, and 38, and four dba⁻ strains 6, 17, 18, and 26, with the Bacterial Outer Membrane Protein Extraction Kit (BestBio, Shanghai, China). The detergent remaining in the extracted OMPs was removed with Amicon Ultra-0.5 10K Centrifugal Filters (Merck Millipore) and the HiPPR Detergent Removal Spin Column Kit (Thermo Fisher Scientific, IL, USA). The results showed that the OMPs extracted from dba⁺ strains 2, 3, and 38 displayed faster hemolysis of duck erythrocytes during overnight incubation on duck blood agar than live R. anatipestifer cells, whereas the OMPs from dba⁺ strain 33, which has relatively weak hemolytic zone on the duck blood agar, and all the tested dba⁻ strains 6, 17, 18, and 26, showed no hemolytic activity on duck blood agar.

In addition, unlike other Flavobacterium species, such as F. psychrophilum , none of the tested dba⁺ and dba⁻ R. anatipestifer strains agglutinated duck, goose, or chicken erythrocytes.

To determine whether the virulence of all the dba⁺ R. anatipestifer strains were higher than that of the dba⁻ strains, the LD₅₀ values of four dba⁺ strains and two dba⁻ strains were determined. One-day-old Cherry Valley ducklings were obtained from the Jinghu Duck Farm (Jiangying, Jiangsu Province, P. R. China) and housed in cages under a controlled temperature of 28–30 °C and a 12 h light/dark cycle, with free access to food and water during the study. Animal care and maintenance were performed in accordance with the Institutional Animal Care and Use Committee guidelines adopted by the Shanghai Veterinary Research Institute. The animal experiment in this study was approved by the Institutional Animal Care and Use Committee of the Shanghai Veterinary Research Institute of the Chinese Academy of Agricultural Sciences (permit no.: 16–10). The LD₅₀ values of dba⁺ strains 3 and 16 in 8-day-old Cherry Valley ducklings were ≥10¹⁰ c.f.u.; those of dba⁺ strains 2 and 33 were 4.80×10⁸ c.f.u. and 1.20×10³ c.f.u., respectively; and those of dba⁻ strains 17 and 18 were 3.81×10⁵ c.f.u. and 6.12×10⁸ c.f.u., respectively (Table 1). Therefore, the highly virulent strain 33 did not have the strongest or weakest hemolytic activity among these six strains, and the virulence of the strong-hemolytic-activity strain 3 was attenuated. These results suggest that a dba⁺ R. anatipestifer strain may not necessarily display higher virulence than a dba⁻ strain, and that the virulence of R. anatipestifer is strain-dependent.

Table 1.

The hemolytic activity and virulence of R. anatipestifer strains

Strain number*	Strain name	Hemolytic activity on the duck blood agar	Hemolytic activity by liquid hemolysis assay (%)†	LD50 to ducklings
2	CH3	dba+	8.00±0.45	4.80×108 c.f.u.
3	NJ-1	dba+	9.03±0.26	≥ 1010 c.f.u.
33	HXb2	dba+	6.85±0.29	1.20×103 c.f.u.
16	SC-12	dba+	5.44±0.59	≥ 1010 c.f.u.
17	Yb2	dba-	4.95±0.23	3.81×105 c.f.u.
18	JY-1	dba-	4.43±0.14	6.12×108 c.f.u.

*Numbers corresponding to the numbers given to the strains in Table S1.

†After bacteria and duck red blood cells incubation in Williams’ Medium E for 24 h, the hemolytic activity of R. anatipestifer strains was detected by liquid hemolysis assay.

In this study, 29 of 52 R . anatipestifer strains showed hemolytic activity on duck blood agar. However, the observation through phase contrast microscopy showed that all the tested dba⁺ and dba⁻ strains in the liquid assay could generate pores in the duck erythrocyte membrane. That the hemolytic activity of dba⁻ strains could not be detected in the blood agar assay may be due to its low sensitivity. Strain 17, just as other dba⁻ R. anatipestifer strains, only generated a few small pores in the duck erythrocytes, but it could grow quickly in blood of the infected ducklings [17]. It suggested that dba⁻ strains could also obtain enough iron from the released hemoglobin and heme, or from other iron sources, to support their growth and proliferation in blood of the infected ducklings.

Our results suggested that both their culture supernatants and OMPs of R. anatipestifer contained hemolysins. In contrast, the cell-free extracellular products of F. psychrophilum , which also belongs to the family Flavobacteriaceae, showed no hemolytic activity [16]. In addition, none of the dba⁺ and dba⁻ R. anatipestifer strains tested in this study agglutinated erythrocytes, while some Flavobacterium species, including F. psychrophilum , F. meningosepticum , and F. indologenes [16], were contact-dependent hemolysis, and they also agglutinated erythrocytes. The hemagglutinating and hemolytic reactions of these Flavobacterium species are known to be dependent on the initial adherence of the bacteria to the erythrocytes and the subsequent lysis of the erythrocytes [16, 18].

Similar to hemolytic F. psychrophilum [16], in this study, the hemolytic capacities of R. anatipestifer strains on duck blood agar did not correspond to their virulence. It suggested that the virulence of R. anatipestifer was strain-dependent, but not hemolysis-related. Although hemolytic capacities of R. anatipestifer strains on duck blood agar did not correspond to their virulence, the distinct septicemia and high blood bacterial loading in R. anatipestifer infected ducks does indicate lysing the duck red blood cells by R. anatipestifer may play a role in the pathogenesis of infection. Moreover, all the tested strains could generate pores in the duck erythrocyte membrane. The exact role of hemolytic activity of R. anatipestifer during bacterial infection and hemolysins involved will be uncovered with further research.