PLoS One. 2017 Jul 25;12(7):e0181942. doi: 10.1371/journal.pone.0181942.

Identification of novel antigen candidates for a tuberculosis vaccine in the adult zebrafish (Danio rerio)


Henna Myllymäki1*, Mirja Niskanen1, Kaisa Ester Oksanen1#b, Eleanor Sherwood1#a, Maarit Ahava1, Mataleena Parikka1,2, Mika Rämet1,3

1 BioMediTech Institute and Faculty of Medicine and Life Sciences, University of Tampere, Tampere, Finland

2 Oral and Maxillofacial Unit, Tampere University Hospital, Tampere, Finland

3 PEDEGO Research Unit, Medical Research Center Oulu, University of Oulu, and Department of Children and Adolescents, Oulu University Hospital, Oulu, Finland



Tuberculosis (TB) remains a major global health challenge and the development of a better vaccine takes center stage in fighting the disease. For this purpose, animal models that are capable of replicating the course of the disease and are suitable for the early-stage screening of vaccine candidates are needed. A Mycobacterium marinum infection in adult zebrafish resembles human TB. Here, we present a pre-clinical screen for a DNA-based tuberculosis vaccine in the adult zebrafish using an M. marinum infection model. We tested 15 antigens representing different types of mycobacterial proteins, including the Resuscitation Promoting factors (Rpf), PE/PPE protein family members, other membrane proteins and metabolic enzymes. The antigens were expressed as GFP fusion proteins, facilitating the validation of their expression in vivo. The efficiency of the antigens was tested against a low-dose intraperitoneal M. marinum infection (≈ 40 colony forming units), which mimics a primary M. tuberculosis infection. While none of the antigens was able to completely prevent a mycobacterial infection, four of them, namely RpfE, PE5_1, PE31 and cdh, led to significantly reduced bacterial burdens at four weeks post infection. Immunization with RpfE also improved the survival of the fish against a high-dose intraperitoneal injection with M. marinum (≈ 10.000 colony forming units), resembling the disseminated form of the disease. This study shows that the M. marinum infection model in adult zebrafish is suitable for the pre-clinical screening of tuberculosis vaccines and presents RpfE as a potential antigen candidate for further studies.

PMID: 28742838



Mycobacterium tuberculosis, the causative agent of TB, caused 1.4 million deaths and 10.4 million new infections in 2015. The only available TB vaccine, Bacillus Calmette Guérin (BCG), protects young children, but its protection in adults is limited. The WHO has set an ambitious goal to eliminate TB as a global health problem by the year 2050 [1]. To reach this, new vaccines that protect from the primary infection, boost BCG-induced immunity or prevent the reactivation of a latent infection, are needed.

For vaccine research, animal models that can replicate the course of the disease and are suitable for the early-stage screening of candidates are essential. However, M. tuberculosis is not a natural pathogen of traditional animal models such as mice, rabbits and guinea pigs. The zebrafish (Danio rerio) with its natural pathogen Mycobacterium marinum has emerged as an advantageous animal to model a TB infection [2].  M. marinum is a close relative to M. tuberculosis, and in zebrafish, an infection leads to a disease that resembles human TB in many aspects including naturally developing latency and reactivation upon immunosuppression [3]. In addition, there are several parallels in immune responses involved in mycobacterial infections in zebrafish and humans, and similar virulence factors are used by both M. marinum and M. tuberculosis [2]. Moreover, we have previously shown that adult zebrafish can be partially protected against mycobacteriosis with the BCG vaccine or with a DNA vaccine expressing a combination of antigens [4].

The current study was designed to test the applicability of the adult zebrafish-M. marinum infection model in the pre-clinical screening of DNA-based tuberculosis vaccines. For this, we selected 15 M. marinum antigens that have a homologue in M. tuberculosis, belong to different functional categories and are expressed during different stages of mycobacterial growth. These included four Resuscitation promoting factors, three PE/PPE family members and five other membrane associated proteins together with three proteins involved in metabolism. Prior to the screening, we verified the expression of the corresponding mycobacterial genes in the M. marinum ATCC 927 strain by qRT-PCR. The antigens were expressed as GFP fusion proteins, which facilitated the confirmation of their expression in vivo by fluorescence microscopy, ELISA and Western blotting. The screening protocol is outlined in Figure 1. The selected antigens were tested as prophylactic (pre-exposure) DNA vaccines using two variations of the zebrafish mycobacterium infection model: a low-dose and a high-dose infection.



Figure 1. Outline of the screening protocol.


In humans, a M. tuberculosis infection most often leads to a latent infection, which is asymptomatic, but can potentially reactivate. In addition to a TB vaccine that would completely prevent new infections, a vaccine that could control the infection and prevent its dissemination would ease the TB problem. In the adult zebrafish, a primary infection can be modelled by a low-dose M. marinum infection (≈40 cfu). In most fish, this leads to a latent disease with stable bacterial counts in approximately five weeks [3]. To assess vaccine efficacy against a low-dose infection, the bacterial burden in each fish was measured by qPCR five weeks post infection. Most of the 15 antigens tested did not affect the bacterial numbers and none of them was able to clear the infection completely. However, four of the candidates significantly reduced the bacterial burden. These included RpfE, which led to an 88% reduction in median bacterial counts; in addition to two PE protein family members, PE5_1 and PE31, and the metabolic protein cdh. These reduced the bacterial burden by 56%, 50% and 62%, respectively (Figure 2.).  This is comparable to the effect of the BCG vaccination in zebrafish [5].

In young children, a M. tuberculosis infection can lead to a fulminant infection, called miliary TB.  The BCG vaccine protects children against this, but its use is limited due to safety issues, for example, HVI-coinfected and immunosuppressed people are excluded [1]. Therefore, there is a need for a safer vaccine for defending infants from the disseminated form of TB. To model a miliary TB infection in the zebrafish, we used a high-dose M. marinum infection (≈10,000 cfu) that leads to an acute disease and relatively high mortality [3]. 10 of the antigens were chosen for assessment in this assay, including the four that reduced the bacterial burden in the low-dose infection assay. Survival of the fish was monitored for 12 weeks, during which all fish showing signs of disease were euthanized. In the high-dose infection assay, the RpfE antigen led to improved survival (40% compared to the 16% of the control group) (Figure 2.).

Safer vaccine alternatives for replacing BCG are being actively investigated. Subunit vaccines, such as DNA vaccines, do not contain live bacteria, but only the antigenic parts of the pathogen. Therefore, they pose no risk of inducing a disease even in immunocompromised people. The antigens for a subunit vaccine are chosen with the intended protective mechanism in mind: a pre-exposure vaccine would contain antigens expressed in replicating M. tuberculosis, while a post-exposure vaccine would consist of antigens expressed during dormancy or reactivation. As several antigens can be displayed in a subunit vaccine, a well-chosen combination would ideally give protection against different stages of TB. In this study, we identified four antigens that reduce the bacterial burden in a primary zebrafish M. marinum infection. Further studies are required to determine their potential as vaccine candidates.

RpfE has been studied to some extent as a DNA vaccine candidate in the mouse model, where it has shown high immunogenicity and variable protection against M. tuberculosis both in terms of cfu burdens and survival times [6]. We have previously shown that a DNA vaccine consisting of RpfE together with two other well-studied antigens, ESAT-6 and Ag85B, can boost the protection provided by the BCG vaccination in the adult zebrafish [5]. This makes the zebrafish also a promising model for developing booster vaccines for BCG.

In conclusion, this study shows that the M. marinum infection model in adult zebrafish is suitable for the pre-clinical screening of tuberculosis vaccines and presents RpfE as a potential antigen candidate for further studies.



Figure 2. RpfE, PE5_1, PE31 and cdh antigens reduce bacterial burden upon a low-dose M. marinum infection and the RpfE antigen also improves survival upon a high-dose infection. (A) Of the 15 vaccine candidates screened, immunization with the RPfE, PE5_1, PE31 or cdh antigens leads to significantly reduced bacterial burden upon a primary M. marinum infection. Adult zebrafish were vaccinated with 12 μg of the experimental or control (GFP) antigen plasmids and infected with a low-dose (~40 cfu) of M. marinum four weeks later. The bacterial burden in each fish was determined five weeks post infection from extracted DNA. N=20-35, each immunization was independently replicated 2-3 times. To enable pooling of several experiments, the number of colony forming units (cfu) detected in each fish was normalized with the median cfu of the GFP-immunized control group of that experiment. The horizontal line represents median, the whiskers show min and max. * p<0.05 (two-tailed Mann-Whitney test) (B) Immunization with the RpfE antigen improves survival upon a high-dose M. marinum infection. The fish were immunized with the RpfE or GFP antigens, followed by infection with a high-dose (~10,000 cfu) of M. marinum four weeks later. The fish were monitored for survival for 12 weeks post-infection (wpi). The fish showing signs of illness were euthanized. N=57 (GFP) and 82 (RpfE). ** p<0.01 (Log-rank (Mantel-Cox) test).



[1] WHO. Global tuberculosis report 2016. 2016;WHO/HTM/TB/2016.13.

[2] Myllymaki H, Bauerlein CA, Ramet M. The Zebrafish Breathes New Life into the Study of Tuberculosis. Front Immunol 2016 May 19;7:196.

[3] Parikka M, Hammaren MM, Harjula SK, Halfpenny NJ, Oksanen KE, Lahtinen MJ, et al. Mycobacterium marinum causes a latent infection that can be reactivated by gamma irradiation in adult zebrafish. PLoS Pathog 2012 Sep;8(9):e1002944.

[4] Oksanen KE, Halfpenny NJ, Sherwood E, Harjula SK, Hammaren MM, Ahava MJ, et al. An adult zebrafish model for preclinical tuberculosis vaccine development. Vaccine 2013 Oct 25;31(45):5202-5209.

[5] Oksanen KE, Myllymaki H, Ahava MJ, Makinen L, Parikka M, Ramet M. DNA vaccination boosts Bacillus Calmette-Guerin protection against mycobacterial infection in zebrafish. Dev Comp Immunol 2016 Jan;54(1):89-96.

[6] Yeremeev VV, Kondratieva TK, Rubakova EI, Petrovskaya SN, Kazarian KA, Telkov MV, et al. Proteins of the Rpf family: immune cell reactivity and vaccination efficacy against tuberculosis in mice. Infect Immun 2003 Aug;71(8):4789-4794.