To the best of our knowledge, this is the first report describing the isolation of a stable cell population releasing fluorescent HIV-1 based VLPs. The biological features of 18-4s cells do not diverge significantly from those of the 293/RGP parental cells , except for a more stringent requirement of sodium butyrate for the VLP production. Sodium butyrate is known acting on methylated promoter sequences and/or on surrounding heterochromatinic DNA, ultimately leading to the re-activation of silenced genes. This is thought occurring through the hyperacetylation of histone and non-histone nuclear proteins [30, 31]. We observed that, in the one hand, the parental 293/RPG cells significantly decreased the levels of VLP release with passages. This was possibly due to a progressive transcriptional silencing of HIV-1 regulatory and/or structural genes, and was efficiently reverted by sodium butyrate. This treatment, on the other hand, was absolutely required for the release of VLPs from 18-4s cells, likely due to the silencing of HIV-1 genes already operative at the time of the cell isolation.
The FACs analysis revealed that 18-4s cells are composed by two clearly distinguishable sub-populations in terms of the expression of both NefG3C-GFP and ΔNGFr. In particular, on the basis of the mean fluorescent intensities we measured by FACs, the bright sub-population seemed to accumulate both products about 5-fold more efficiently than the dull one. Concerning the most relevant expression of NefG3C-GFP, we noticed that with passages, the bright sub-population tended to reduce, meanwhile appearing a detectable fraction of GFP negative cells. Again, this is likely the consequence of the silencing of the NefG3C-GFP sequences rather than to their loss since also in this case the treatment with sodium butyrate resulted in the restoration of the original GFP profile (not shown). We have no obvious explanations for the mechanisms underlying the production of the two sharply distinct 18-4s cell sub-populations, even if it may be conceivable that this reflects the differences in the number and/or sites of the integration of the plasmid vectors. In addition, the consequent protein overproduction might be associated with an impaired protein degradation efficiency, thus resulting in steady-state protein levels overriding the real differences in the numbers of integrated vectors.
The 2–3 fold decreased levels of VLP release from the 18-4s cell line compared with the parental cells appeared consistent with the reduced viral production we observed in transient transfection experiment (not shown), and was possibly due to some steric hindrance occurring during the assembling/release process as the consequence of the high levels of NefG3C-GFP incorporation. Of note, the data from EM analysis showed that this did not lead to significant morphologic alterations of VLPs that in both cases appeared as both mature and immature particles. In this regard, we cannot formally exclude that the significant presence of immature viral particles depended at least in part on a slow kinetic of maturation rendering more difficult the detection of mature particles through the electron microscope observation of ultra-thin cell sections we performed.
The analyses we carried out by FACs and by fluorescence and confocal microscopes consistently indicated that the cells targeted by pseudotyped 18-4s VLPs become fluorescent as early as 2 hours after the challenge. Generally speaking, the great part of HIV-1 particles undergo cell internalization through a receptor independent endocytosis leading to lysosome degradation [10, 32], whose inhibition, in fact, increases HIV infection [33, 34]. In the analysis of the VLP entry we performed, it is worthy of note that the 18-4s VLPs pseudotyped with the fusion defective VSV-G or with "null" VLPs did not induce a significant increase of the cell fluorescence compared with the mock challenged cells. These results, together with those obtained with chemical inhibitors or neutralizing antibodies, strongly suggests that the cell fluorescence we detected basically relied on the completion of the virus fusion events with the release into the cytoplasm of the fluorescent molecules, a conclusion also supported by our results from the confocal microscope analysis.
We feel the relevance of the 18-4s cells stands in two main fields of application. The first could be the study of the mechanisms and dynamic of viral entry, and the second regards the screening of antiviral compounds targeting the late HIV replication events. In particular, the possibility to pseudotype HIV-1 VLPs with a wide array of viral receptors could allow gaining more insights on the not yet fully elucidated mechanism of entry of viruses of different species. This is for instance the case of Human hepatitis C virus, whose cell receptor/co-receptors have still to be unambiguously identified (for a review, see ). In this regard, the cell fluorescence generated by the VLP entry would be a more stringent marker of viral entry compared with the in use systems relying on the lentivirus-mediated transduction of gene markers whose expression can be restricted by host cell factors. On the other hand, the rather simple protocol for the detection of fluorescent VLPs by FACs we developed renders 18-4s cells a both safe and powerful tool for large-scale screenings of new anti-HIV compounds electively targeted to the assembly/release processes. Thus, 18-4s cells are anticipated to be a useful reagent for multiple purposes that will be freely available upon request.
Furthermore, from the point of view of the basic virology, the possibility to analyze viral particles by FACs as here described opens the way towards new rapid and accurate approaches in the investigation on the molecular composition of enveloped viruses in terms of both viral and virus-associated cell products.
Finally, on the basis of the results we obtained with the MLV-based VLPs, we expect that the NefG3C-GFP approach could be applied also to different retroviral systems close to HIV-1(e.g., HIV-2, SIV) and MLV.