Of nanoparticles, cells and polyanions
It is the end of semester 2 so it’s marking season. Since we double mark (a good thing), the final year research projects are marked by both supervisor and an assessor, a member of staff who is not involved in the project. One of the projects I marked was Gemma Carolan’s on “How do SmartFlares RNA detection probes reach the cytosol? Available are the PDF of report, and posts here and here.
I had a sense of déjà vu while reading the project – the clear endosomal location of the SmartFlares, regardless of the DNA sequences brought me back to the days when antisense was the technology of the future for medicine.
While evaluating new technology it is useful to go back and look at other high flying technology. The reality is that it takes decades before we know whether the promise (and hype) were justified; this is true for any hot topic from stem cells to nanoparticles and graphene.
Antisense effects can be mediated by RNAse H, an enzyme that specifically cleaves RNA-DNA duplexes and which protects our cells from RNA viruses. There are other mechanisms, e.g., interference with splicing or translation, but the RNAse-H mediated transcript degradation should be central to many antisense effects. There were many papers reporting specific effects (evidenced by differences between sense, antisense and scrambled oligonucleotides sequences). These certainly contributed to success of individuals and of institutions, e.g., in UK Research Assessment Exercise and grant awards.
Many were happy with what is in retrospect a superficial level of evidence. However, there was a group of ‘discontents’, who wanted mechanistic evidence that what was observed at the biological level was indeed due to the proposed mechanism. At this point some of the wheels began to come off the antisense bus. A series of papers demonstrated that antisense oligonucleotides have substantial non-antisense effects, in cultured cells and in vivo: some examples are vasculature smooth muscle cell proliferation here, here, and the very elegant demonstration that bcl-2 targeted ‘antisense’ oligonucleotides have the same effect when bcl2 mRNA is down regulated by other means, illustrates how the activity of the antisense oligonucleotide is clearly due to off-target activity and not due to interfering with bcl2 expression
Oligonucleotides are polyanions and the evidence for protein interactions in the extracellular space with sulfated glycosaminoglycans controlling protein and cell function was nascent, but growing. Evidence accumulated that competing with endogenous heparan sulfate (one of the key extracellular polyanions) was one non-antisense mechanism of “antisense” oligonucleotides. Moreover, the oligonucleotides demonstrably bound to heparin- and HS-binding extracellular proteins and that such effects could be sequence-dependent, could replace the heparan sulfate co-receptor for fibroblast growth factors, and by interacting with a heparin-binding integrin, Mac-1 (CD11b/CD18; alpha M beta 2), could exert profound effects on cell adhesion.
It is important to distinguish mechanism from clinical potential, as this paper shows that the general polyanionic effect of ‘antisense’ oligonucleotides may have benefits, though the mechanism has nothing to do with base pairing.
The reason I am aware of some of this work, is that in Liverpool an antisense group led by David Tidd, demonstrated conclusively that RNAse-H mediated transcript cleavage could be achieved with exquisite specificity BUT ONLY IF the antisense oligonucleotides were physically put into the cytosol, in this instance by transient cell permeablisation with streptolysin-O. This paper was, I think, the first to demonstrate such a mechanism in mammalian cells. They subsequently published an elegant comparison of localisation of oligonucleotides in Blood, which is well worth a read by the entire community involved in trying to target entities to the nucleocytoplasmic space from outside the cell. This paper sets out clearly what needs to be done to demonstrate such a phenomenon.
In general, oligonucleotides added to cells simply bound to the pericellular matrix, and, after interacting with cell surface proteins, were taken up into the endosomal system. The extracellular interactions were one route whereby oligonucleotides could elicit changes in the functions of cells. Since the current list of extracellular proteins that bind to heparan sulfate currently stands at 435 the number of potential targets for polyanions such as oligonucleotides is clearly substantial.
Gemma’s work on SmartFlares, which can be accessed via her supervisor’s blog, does have some parallels with the world of antisense two decades ago. The SmartFlares are polyanions. Protein recognition of extracellular polyanions (in vivo, glycosaminoglycans such as heparan sulfate) is dependent on charge, but is also selective, so the disposition of charges in space determines whether proteins will bind or not (posts on some papers I have been involved in here, here, and here).
Single stranded DNA has a complex, sequence-specific structure(s) and double stranded DNA is not 100% B-DNA. Therefore, off target sequence specific effects should not surprise anyone. Gemma’s work demonstrates clearly endosomal localisation and an absence of detectable cytosolic localisation of the SmartFlares.
So have we come full circle? Only in part; we are not exactly on the same track. We have moved from antisense to nanoparticle-nucleic acid conjugates, so we might want to add to Langmuir’s concept of ‘pathological science’ a dose of re-incarnation for the “Smart Flares”. However, there is growing recognition that nanoparticle conjugates do not enter the nucleocytoplasmic space and indeed the most recent paper on the nucleic acid-nanoparticle conjugates does not claim anything more than endosomal localisation.
The latter paper also provides an illustration that such engineered polyanions can have profound effects on cells, in this instance on the immune system. This is hardly surprising given the previous work on ‘antisense’ oligonucleotides and the fact that many of the molecular systems that regulate the immune system are themselves regulated by binding to heparan sulfate, from cytokines to adhesion receptors.
In conclusion, the evidence for ‘Smart Flares” acting via Watson-Crick bonding to nucleocytoplasmic nucleic acids is absent, simply because we do not have strong evidence for their activity in the nucleo-cytoplasmic space. In contrast, engineering polyanions on nanoparticles to present novel interacting surfaces to extracellular regulatory proteins is the latest in a long line of work, which goes all the way back to heparin, which by weight, is the biggest drug in the world, with a market value (heparin, derivatives, etc.) of ~$5 billion. So the future may well be bright for these materials, as modulators of cell communication, acting on the network of proteins that bind sulfated glycosaminoglycans.