Background Mosquito species from the Anopheles gambiae complex and the Anopheles funestus group are dominant African malaria vectors. Mosquito microbiota play vital roles in physiology and vector competence. Recent research has focused on investigating the mosquito microbiota, especially in wild populations. Wild mosquitoes are preserved and transported to a laboratory for analyses. Thus far, microbial characterization post-preservation has been investigated in only Aedes vexans and Culex pipiens. Investigating the efficacy of cost-effective preservatives has also been limited to AllProtect reagent, ethanol and nucleic acid preservation buffer. This study characterized the microbiota of African Anopheles vectors: Anopheles arabiensis (member of the An. gambiae complex) and An. funestus (member of the An. funestus group), preserved on silica desiccant and RNAlater® solution. Methods Microbial composition and diversity were characterized using culture-dependent (midgut dissections, culturomics, MALDI-TOF MS) and culture-independent techniques (abdominal dissections, DNA extraction, next-generation sequencing) from laboratory (colonized) and field-collected mosquitoes. Colonized mosquitoes were either fresh (non-preserved) or preserved for 4 and 12 weeks on silica or in RNAlater®. Microbiota were also characterized from field-collected An. arabiensis preserved on silica for 8, 12 and 16 weeks. Results Elizabethkingia anophelis and Serratia oryzae were common between both vector species, while Enterobacter cloacae and Staphylococcus epidermidis were specific to females and males, respectively. Microbial diversity was not influenced by sex, condition (fresh or preserved), preservative, or preservation time-period; however, the type of bacterial identification technique affected all microbial diversity indices. Conclusions This study broadly characterized the microbiota of An. arabiensis and An. funestus. Silica- and RNAlater®-preservation were appropriate when paired with culture-dependent and culture-independent techniques, respectively. These results broaden the selection of cost-effective methods available for handling vector samples for downstream microbial analyses.

The hypomethylating agent 5-azacytidine (AZA) is the first-line therapy for acute myeloid leukemia (AML) patients unfit for intensive chemotherapy. Evidence suggests that the anti-tumor effect of AZA results partly from T-cell cytotoxic responses against MHC-I-associated peptides (MAPs) whose expression is induced by hypomethylation. Through a proteogenomic approach, we analyzed the impact of AZA on the transcriptome and MAP repertoire of four AML cell lines and validated salient findings in the transcriptome of 437 primary AML samples. We demonstrate that AZA caused pleiotropic changes in AML cells via perturbation of transcription, translation, and protein degradation. Overall, 1,364 MAPs were upregulated in AZA-treated cells, including several cancer-testis antigens. Increased MAP abundance was due to the upregulation of corresponding transcripts in a minority of cases and post-translational events in most cases. Furthermore, AZA-induced hypomethylation increased the abundance of numerous transcripts, of which 38% were endogenous retroelements (EREs). Upregulated ERE transcripts triggered innate immune responses but were degraded by autophagy and not processed into MAPs. Autophagy resulted from the formation of protein aggregates caused by AZA-dependent inhibition of DNMT2, a tRNA-methyl transferase enzyme. We found that autophagy inhibition had a synergistic effect with AZA on AML cell proliferation and survival, increased ERE levels and triggered pro-inflammatory responses. Finally, autophagy gene signatures were associated with a lower abundance of CD8+ T-cell markers in AML patients expressing high levels of EREs. Altogether, this work demonstrates that the impact of AZA is regulated at several levels and suggests that inhibiting autophagy could improve the immune recognition of AML blasts in patients.