Rickettsia conorii infection stimulates the expression of ISG15 and ISG15 protease UBP43 in human microvascular endothelial cells

Abstract

Rickettsia conorii, an obligate intracellular bacterium and the causative agent of Mediterranean spotted fever, preferentially infects microvascular endothelial cells of the mammalian hosts leading to onset of innate immune responses, characterized by the activation of intracellular signaling mechanisms, release of pro-inflammatory cytokines and chemokines, and killing of intracellular rickettsiae. Our recent studies have shown that interferon (IFN)-β, a cytokine traditionally considered to be involved in antiviral immunity, plays an important role in the autocrine/paracrine regulation of host defense mechanisms and control of R. conorii growth in the host endothelial cells. Here, we show that R. conorii infection induces the expression of ISG15 (an interferon-stimulated gene coding a protein of 17kD) and UBP43 (an ISG15-specific protease) at the levels of mRNA and protein and report the evidence of ISGylation of as yet unidentified target proteins in cultured human microvascular endothelium. Infection-induced expression of ISG15 and UBP43 requires intracellular replication of rickettsiae and production of IFN-β, because treatment with tetracycline and presence of an antibody capable of neutralizing IFN-β activity resulted in near complete attenuation of both responses. Inhibition of R. conorii-induced ISG15 by RNA interference results in significant increase in the extent of rickettsial replication, whereas UBP43 knockdown yields a reciprocal inhibitory effect. In tandem, these results demonstrate the stimulation of interferon-β-mediated innate immune mechanisms capable of perturbing the growth and replication of pathogenic rickettsiae and provide first evidence for ISG15-mediated post-translational modification of host cellular proteins during infection with an intracellular bacterium.

Figure 1

Activation of ISG15 and UBP43 mRNA expression during R. conorii infection of microvascular endothelium in vitro. Total RNA collected from HMECs infected with R. conorii (Rc) for 48 and 72 h was subjected to quantitative real-time PCR to determine the steady-state levels of transcripts for ISG15 and UBP43 using glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as a housekeeping gene. Changes in the contents of ISG15 (A) and UBP43 (B) are shown as the mean ± standard error of the mean from a minimum of three independent experiments. For the ease of comparison, the basal mRNA levels of each gene in uninfected cells (Control) were given a value of 1.

Figure 2

R. conorii infection of HMECs up-regulates ISG15 and UBP43 protein expression and ISGylation. A. Western blot anaIysis of ISG15 and UBP43 in HMECs infected with R. conorii for different durations of time (24, 48, and 72 h). The antibody for UBP43 was capable of detecting both full length (39kDa) and amino-terminal truncated (34kDa) isoforms of UBP43. Also shown in the effect of tetracycline (TC) treatment on the levels of ISG15 and UBP43 expression during R. conorii infection. The results of a typical experiment (n=3) are presented. The relative intensity of each band in comparison to the level of baseline expression in uninfected controls is also shown. B. Evidence for protein ISGylation during R. conorii infection and IFN-β treatment of HMECs. Protein extracts were prepared from HMECs infected with R. conorii, treated with recombinant human IFN-β, or those left uninfected and untreated for controls. Equal amounts of protein were separated on an SDS-PAGE gel and electro-transferred to a nitrocellulose membrane. The membrane was then probed with an anti-ISG15 antibody. The arrow indicates a non-specific, cross-reactive protein band present in all experimental conditions, the intensity of which is variable even among simultaneously processed controls. The levels of free ISG15 in response IFN-β treatment and infection in the presence and absence of tetracycline and the effect of inhibition of rickettsial replication on ISGylation are also presented.

Figure 3

IFN-β secreted by R. conorii-infected endothelial cells induces ISG15 and UBP43 expression. An antibody capable of neutralizing the activity of IFNβ was added to the culture medium during the infection. Total protein lysates were prepared at different times post-infection and processed for the detection of ISG15 and UBP43 by western blotting. The relative intensity of each band in comparison to the level of baseline expression in uninfected controls at 72 h, which was assigned a value of 1, was determined by densitometry and normalization to the housekeeping protein α-tubulin.

Figure 4

Effect of ISG15 and UBP43 knockdown on R. conorii replication in HMECs. HMECs were transfected with ISG15 and UBP43-specific siRNA or non-target control siRNA prior to infection with R. conorii. To ensure the efficacy of knockdown, the levels of ISG15 (A) and UBP43 (B) were first investigated by Western blot analysis. Next, the levels of intracellular replication of R. conorii in HMECs transfected with siRNAs against ISG15 and UBP43 were determined by rickettsial outer membrane protein A (rOMPA)-based quantitative PCR. Rickettsial replication was significantly increased in ISG15-knockdown cells (C), but was significantly lower than the corresponding controls in UBP43-siRNA transfected cells (D).