Rickettsiae are Gram-negative alpha proteobacteria and arthropod-borne disease agents of spotted fevers and typhus [1]. Because of their obligate intracellular lifestyle, the molecular mechanisms involved in their pathogenicity are still poorly understood. To circumvent the difficulties in working with these bacteria, rickettsial genomes have been sequenced, thus allowing further postgenomic studies. DNA microarrays can be considered as powerful tools to understand the host–pathogen interactions in the course of infectious diseases. Exploring the RNA profiles of both host and pathogen promises to improve our knowledge of the infectious process, thus favouring the development of preventive or therapeutic strategies [2]. While published studies mainly focused on the eukaryotic response, the complementary picture, i.e. the bacterial response, was poorly analysed because of several technical limitations [3]. The transcriptomic analysis of R. conorii was recently performed by microarrays [4]. This bacterium is responsible for Mediterranean spotted fever (MSF), a disease characterised by a skin lesion called eschar caused by the bite of its vector, the brown dog tick, Rhipicephalus sanguineus [5]. The aim of this work was to analyse the transcriptome pattern of rickettsiae within such eschars using the same strategy based on the removal of eukaryotic contaminants coupled with subsequent random amplification of cDNA. Our goal was also to recover eukaryotic RNA from the same sample, thus considering both sides of the host–pathogen interaction. Preliminary assays were performed using rabbit eschars generated by the sub-cutaneous inoculation of purified bacteria (not shown). Human eschars were collected by sterile scalpel excision from patients suffering from MSF and immediately placed at 4 C overnight in 1 mL of RNA Later (Qiagen, Valencia, California, USA) before storage at )80 C. The biopsies were rapidly decontaminated by a 5-min incubation in iodated alcohol and 1 min washing in nuclease free water. The tissues excised in small pieces were mechanically lysed with tungsten beads and using the Mixer Mill MM3 (Qiagen), before enzymatic digestion with proteinase K. The total RNA was extracted from resulting lysates using the RNeasy Micro kit (Qiagen). Eukaryotic RNA was removed from the total RNA sample using the MicroEnrich Kit (Ambion, Austin, Texas, USA) as described [4]. To recover the eukaryotic RNA from the same sample, the beads were then mixed in TE buffer pH 8 and heated for 1 min at 95 C. The supernatants were collected and precipitated with alcohol. The bacterial RNA was retro-transcribed using random primers and amplified with the Genomiphi kit (GE Healthcare, Piscataway, New Jersey, USA), while human RNA was amplified using poly d(T) primers with T7 polymerase (Agilent, Santa Clara, California, USA). To check for the quality of amplification, PCR were performed on cDNA using primers targeting the rrs gene from R. conorii (FGGCTCAGAACGAACGCTATC ⁄R-GTTAGCTGCGAAACCGAAAG) and the eukaryotic actin gene (F-GGACTTCGAGCAAGAGATGG ⁄R-AGCACTGTGTTGGCGTACAG), respectively. Hybridisations of prokaryotic cDNA were achieved following Cy3 or Cy5 labelling of R. conorii from eschars and using bacteria grown in vero cells as control. R. conorii microarrays were obtained from Agilent (‘Custom microarray’ design) and processed as described [6]. Hybridisations of human microarrays (Agilent) were performed with one colour labelling as recommended. The electrophoregrams obtained with the Bioanalyzer (Agilent) indicated that the extraction of Corresponding author and reprint requests: P. Renesto, Universite de la Mediterranee, URMITE IRD-CNRS 6236, Faculte de Medecine, 27 Bd Jean Moulin, 13385 Marseille, cedex 05, France E-mail: patricia.renesto@univmed.fr
Read full abstract