A repeat motif on a Coxiella effector protein facilitates apoptosis inhibition

KEYWORDS:

• apoptosis inhibition
• Coxiella
• effector proteins

This article refers to:

The inhibition of the apoptosis pathway by the Coxiella burnetii effector protein CaeA requires the EK repetition motif, but is independent of survivin

Intracellular pathogens manipulate host cells in myriad ways. Coxiella burnetii, a slow-growing (generation time ˜11 h), obligately intracellular bacterial pathogen inhibits host cell apoptosis in order to generate a replicative vacuole. This subversion is critical to its pathogenicity; however, very little mechanistic details are known about how the pathogen usurps the host cell's regulatory machinery. It is imperative that we gain better insights into Coxiella's pathogenicity because the select agent causes human Q fever and chronic endocarditis.¹ C. burnetii is found worldwide, and a recent outbreak in Netherlands involved thousands of human cases and resulted in the culling of a large number of goats (a) primary reservoir.² Additionally, untreated chronic Q fever, usually presented as endocarditis, is associated with high mortality rates.¹

Apoptosis occurs through a complex cascade of events that involves both extrinsic and intrinsic pathways. The extrinsic pathway acts through the cell surface death receptor signaling, whereas the intrinsic pathway is activated through the mitochondrial membrane proteins.³ Coxiella is known to manipulate multiple factors along both the extrinsic and intrinsic apoptosis pathways.⁴ For instance, Coxiella activates anti-apoptosis kinases AKT, ERK1/2 and PKA that prevent the onset of apoptosis, and induces the expression of several apoptosis-related microRNAs.^5-8 Additionally, 3 effector proteins (AnkG, CaeA and CaeB) secreted through Coxiella's Type IV Secretion System (T4SS) inhibit apoptosis, although their mechanisms of action are not fully understood.^9,10 AnkG was shown to interact with the mitochondrial protein p32, indicating that it inhibits intrinsic apoptosis by decreasing cytochrome c release. Intriguingly, this protein (along with p32) is translocated to the host cell nucleus for as yet unknown functions.^9,10 CaeA and CaeB inhibit apoptosis at the mitochondrial level but in a p32-independent fashion.¹¹ In this issue, Bisle et al. further clarifies the mechanism of action of CaeA. They show that the effector protein inhibits both extrinsic and intrinsic apoptotic pathways, and reveal that its anti-apoptotic activity is dependent on an EK (glutamic acid, lysine) repeat motif.¹²

Throughout its prolonged infectious cycle, C. burnetii continuously manipulates host processes, including host cell death, thus creating a stable intracellular niche.³ This is an important virulence property that allows this obligate intracellular pathogen to maintain host viability despite inducing stress that would normally activate apoptosis.¹³ Coxiella encodes a T4SS that is essential for intracellular replication,^14,15 and over a hundred effector proteins secreted through this apparatus have been identified (for e.g.,^16-21). However, the functions of most of these substrates are not understood. Legionella pneumophila, a close relative, also encodes a similar secretion system, but only a few T4SS substrates are shared between the 2 pathogens, making it difficult to use a comparative approach to understand the functions of effector proteins secreted by Coxiella into the host cell.^9,22 Similarly, different strains of C. burnetii encode diverse repertoires of effector proteins,²³ indicating that T4SS substrates evolve rapidly in a species- or even strain-specific manner. In accord with this view, Bisle et al. show that caeA from 25 strains of C. burnetii contain variable number of EK repetition motif, and that the number of EK repeats embedded within the coiled-coil domain of CaeA appears to have an effect on apoptosis inhibition.¹²

Because C. burnetii is a highly infective pathogen that requires a Biosafety Level-3 facility for conducting experiments, most studies have been performed using the Nine Mile Phase II (NMII) strain, which can be handled under Biosafety Level-2 containment. But the genome of the NMII strain has not yet been sequenced, and hence most researchers utilize the published genome of Nine Mile Phase I strain (NMI), which sometimes lead to inconclusive results. For instance, caeA (CBU_1524) is not annotated as a protein-coding gene in the current version of the NMI genome (NC_002971.3), instead, it is now part of the pseudogene CBU_1523. To check whether the putative pseudogene is transcribed, I analyzed a transcriptome dataset generated from Vero cells infected with C. burnetii NMII.²⁴ As shown in Figure 1, CBU_1523 is indeed expressed during intracellular growth, and appears to contain 2 separate transcription start sites that correspond, respectively, to a small open reading frame of unknown function, and to caeA. Additionally, both proteins seem to have important infection-specific roles because they are expressed 5–10 fold higher intracellularly than in ACCM2 medium (Fig. 1). The RNA-seq data also suggests that several other presumed pseudogenes in NMII could be functional, thereby explaining —at least partially—why transposon insertions in several presumed pseudogenes resulted in reduced intracellular growth.²⁵ Based on these observations, it is clear that a fully annotated genome of the laboratory workhorse NMII strain is urgently required. Having access to the correct genome sequence will further the tremendous progress the research community has already made in comprehending Coxiella's biology and virulence using genome-scale methodologies (for e.g.,^26-30).

Figure 1. Expression patterns of the annotated pseudogene CBU_1523 (gray wide arrow) in C. burnetii grown in the cell-free medium ACCM2 (green line), and in Vero cells (red line) for 72 hours. Two apparent transcription start sites (TSSs) that possibly correspond to a small open reading frame (sORF) and to caeA are indicated by black thin arrows.

Although it is clear that CaeA traverses to host cell nucleus and blocks apoptosis, the complete molecular details of this process are yet to be worked out. Further studies are required to identify the proteins that possibly interact with CaeA, and to determine whether the interacting proteins are required for nuclear localization and/or for inhibiting apoptosis. Furthermore, the recent development of a genetic system,³¹ should enable the generation of caeA-deletion strains of C. burnetii, and to directly analyze the importance of this protein to Coxiella-mediated inhibition of apoptosis. This is especially important because, as noted by the authors,¹² the orthologs of caeA in a majority of Coxiella strains seem to contain frame-shift mutations that would render CaeA non-functional. Collectively, the approaches used by Luhrmann and colleagues^9-13 should serve as a model for future studies aimed at uncovering the functions of all effector proteins secreted by C. burnetii. Because of the lack of conservation of Coxiella effector proteins both within and between species, this will be a slow and painstaking process, which could be accelerated using genome- and transcriptome-enabled approaches. This is an important endeavor that will enlighten us about the distinct biology of an enigmatic pathogen that has uniquely figured out how to make a living in an inhospitable lysosome-derived intracellular compartment.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.