Use of Bacillus thuringiensis Cry3Aa to deliver therapeutic proteins to cells

This week's featured article:

Yang Z, Lee MM, Chan MK. Efficient intracellular delivery of p53 protein by engineered protein crystals restores tumor suppressing function in vivo. Biomaterials. 2021 Mar 16:120759. DOI: 10.1016/j.biomaterials.2021.120759

Bacillus thuringiensis (Bt) stands out among its close relatives by a distinguishing characteristic: the formation of large, regular protein crystals alongside the developing endospore during stationary phase. The crystals can take a wide variety of shapes, depending on which members of the crystal protein family (Cry proteins) comprise them (see Adalat et al. for a recent review). The focus of today's paper, the Cry3Aa protein, spontaneously assembles into a rod-shaped rhomboid crystal in vivo. When the Bt mother cell lyses to liberate the mature spore, the crystal is also released into the environment. The crystals are remarkably stable. Eventually they may be ingested by a target larva from a narrow range of invertebrate host species. They dissolve in the larval digestive tract and eventually insert into the midgut epithelial cells, leading to pore formation, tissue damage, and growth arrest or death.

That is the usual narrative for Cry proteins, and it's easy to see why: these proteins are the basis of an enormously successful alternative to organic chemical insecticides in pest management. In this week's #BacillusJournalClub, however, we will examine another potential application, the use of Cry3Aa as a delivery vehicle for therapeutic proteins to human cells.

This prospect is attractive, in part because Bt synthesizes such massive quantities of Cry proteins. In their classic 1995 minireview, Agaiise and Lereclus asked, "How does Bacillus thuringiensis produce so much insecticidal crystal protein?" They pointed out that the crystal can account for up to 25% of the dry weight of sporulated cells, implying that each cell must synthesize a least a million Cry protein molecules during stationary phase. The answer to their question is fairly well understood for Cry3Aa, although the details are beyond the scope of this discussion. But we can consider a seemingly straightforward application: What if we fused a cargo protein to Cry3A and allowed the Bt host cell to make a million copies, package them into compact, resilient crystals, and spontaneously release them for our use?

The remaining piece to the puzzle is how to deliver the crystals to the cytosol of a human cell. In invertebrate larvae, Cry3Aa requires a receptor protein. It can't simply penetrate the plasma membrane of a human cell, the way it might an epithelial cell in a beetle larva, for example. Instead, it must first be taken up by endocytosis. Then it must somehow escape from the membrane-bound endosome to deliver its cargo intracellularly. The required mechanism would resemble that of bacterial toxins such as botulinum neurotoxin, rather than that of a typical Cry protein. (For a thorough explanation of endosome escape by bacterial toxins, see Williams and Tsai, 2016).

This is the challenge addressed by members of the Michael Chan lab at The Chinese University of Hong Kong (Yang et al. 2021). As pointed out in the paper's introduction, 7 of the top 10 drugs sold worldwide are monoclonal antibodies and are therefore protein-based (see Urquhart L. 2020). But these therapeutics are limited to extracellular use. The Chan lab's ultimate goal was to develop a cytoplasmic delivery system for the p53 tutor suppressor protein, a promising anti-cancer therapeutic. The lab had previously shown that Cry3Aa crystals could readily enter phagocytic macrophages by endocytosis, but that they were subsequently entrapped in endosomes (Yang et al. 20190). Further, the Cry3Aa delivery protein was taken up very poorly by other cell types.

To address these limitations, the authors designed a Cry3Aa derivative with an increased number of positively charged surface residues, since lysine and arginine are known to be important for mammalian cell uptake of charged proteins. Their mutant, now with 11 additional surface exposed lysines, they termed Pos3Aa. Importantly, they demonstrated that Pos3Aa formed crystals of a size and shape virtually indistinguishable from native Cry3Aa when synthesized in Bt hosts. Stability studies and X-ray diffraction structure determination confirmed that no large-scale changes had occurred in Pos3Aa. Would this derivative be more successful at delivering a cargo protein into the cytoplasm of a variety of cell types?

Several experiments show that the answer is yes. First, fluorescently labeled Pos3Aa was taken up readily into a variety of non-phagocytic cell types. As expected, native Cry3Aa was not. Temperatures and inhibitors known to block endocytosis prevented uptake of Pos3Aa. Crucially, imaging experiments demonstrated that much of the labeled Pos3Aa was released from the endosomes and localized in the cytoplasm.

In a proof of concept experiment, the fluorescent mCherry protein was fused to the C-terminus of Pos3Aa, and the fusion proteins, when produced in Bt hosts, retained the size and shape characteristic of Cry3Aa. Pos3Aa-mCherry fusion proteins were taken up into A549 cells with extremely high efficiency, with significant levels of delivery to the cytoplasm.

Encouraged by these results, the authors fused Pos3Aa-mCherry to p53, a transcription factor that is activated with oncogenic stress to regulate expression of several cell cycle, DNA repair, and apoptosis genes that function in the suppression of cancer. The authors point out that nearly half of human cancers are associated with loss of p53 function (see Lacroix et al. 2006). Importantly, p53 fusions were taken up by cells much more efficiently when joined to the Pos3Aa carrier than when presented to the cells as free protein. Pos3Aa-p53 was observed to be taken up by cells from a breast-cancer cell line deficient in p53; the fusion protein escaped from the endosomes and accumulated in the nucleus, where a transcription factor would function. Unexpectedly, then, it appeared that the cargo protein was being released from the fusion intracellularly. Pos3Aa-p53 treatment inhibited growth of the cells in a dose-dependent manner, and images showed membrane blebbing, as occurs during apoptosis. Treated cells also showed a significant level of arrest in the G1 phase of the cell cycle. Finally, the authors demonstrated that pre-treatment of the breast cancer cell line with Pos3Aa-p53 sensitized them to the chemotherapeutic drug 5-fluorouracil. When mice bearing tumors established from the cell line were treated with a combination of Pos3Aa-p53 and 5-fluorouracil, they showed a measurably greater reduction in tumor growth than observed in mice treated with the protein or the drug alone.

These exciting results require verification and further optimization. But it appears that modified Cry3Aa proteins may indeed be useful carriers for delivery of therapeutic proteins to mammalian cells.

Other cited references:

Adalat R, Saleem F, Crickmore N, Naz S, Shakoori AR. In Vivo Crystallization of Three-Domain Cry Toxins. Toxins (Basel). 2017 Mar 9;9(3):80. doi: 10.3390/toxins9030080. PMID: 28282927; PMCID: PMC5371835.

Agaisse H, Lereclus D. How does Bacillus thuringiensis produce so much insecticidal crystal protein? J Bacteriol. 1995 Nov;177(21):6027-32. doi: 10.1128/jb.177.21.6027-6032.1995. PMID: 7592363; PMCID: PMC177438.

Lacroix M, Toillon RA, Leclercq G. p53 and breast cancer, an update. Endocrine-related cancer. 2006 Jun 1;13(2):293-325.

Urquhart L. Top companies and drugs by sales in 2019. Nat Rev Drug Discov. 2020 Apr;19(4):228. doi: 10.1038/d41573-020-00047-7. PMID: 32203287.

Williams JM, Tsai B. Intracellular trafficking of bacterial toxins. Curr Opin Cell Biol. 2016 Aug;41:51-6. doi: 10.1016/j.ceb.2016.03.019. Epub 2016 Apr 13. PMID: 27084982; PMCID: PMC4983527.

Yang Z, Zheng J, Chan CF, Wong IL, Heater BS, Chow LM, Lee MM, Chan MK. Targeted delivery of antimicrobial peptide by Cry protein crystal to treat intramacrophage infection. Biomaterials. 2019 Oct 1;217:119286.