It has been shown in the literature that inhibition of 5α-reductase (5αR) can reduce testosterone’s conversion to dihydrotestosterone (DHT). Excess DHT shortens anagen and causes hair follicles to enter catagen in advance, resulting in hair loss. [1] When activated by mechanosensory signals, Piezo1 can cause hair follicle stem cells (HFSC) depletion, which is often observed in AGA. [2] Wnt/β-catenin signaling pathway is associated with hair regeneration.[3]
Therefore, we designed and constructed 14 plasmids, namely pcDNA3.1–siRNA(5αR)(1-6)-mCherry, pcDNA3.1–siRNA(Piezo1)(1-6)-mCherry, pcDNA3.1-Box CD mini-beta catenin-mCherry and pcDNA3.1-CD63-L7Ae-mCherry. For siRNA encoding plasmids, we transfected them into RM-1 cells to test their efficacy. Then we transfected the most efficient into 293T cells to gain therapeutic sEVs.
Theoretically, these therapeutic sEVs can reach damaged follicles and release the contained 5αR-siRNA, Piezo1-siRNA, and mRNA (β-catenin). Then DHT will be dramatically decreased and the activation of Piezo1 will be cut off. β-catenin will activate the Wnt pathway to promote the proliferation of follicles.
With the above three strategies, hair growth can be promoted.
Besides, to facilitate the translational application of our project, we adopted a scalable sEV purification method—tangential flow filtration and bind-elute size exclusion chromatography (TFF/BE-SEC). After comparing it with conventional methods (UC and EIK), we demonstrated TFF/BE-SEC is the most advantageous one.
Our project involved three cell lines, namely HEK-293T, RM-1 and dermal papilla cells (DPC). HEK-293T, human embryonic kidney cell line, which secretes high levels of sEVs, was used for sEVs production.
RM-1, mouse prostate cancer cell line, which expresses the two target genes (5αR and Piezo1) were used to screen effective siRNAs.
Plasmids expressing siRNA of 5αR and Piezo1 were transfected into RM-1 respectively. The designed plasmids were loaded with mCherry fluorescent protein, so the presence of fluorescence could determine whether the plasmids were successfully transfected into cells. Then we conducted RT-qPCR and WB to determine whether the expression level of target genes was reduced.
After siRNA screening, we transfected plasmids expressing effective siRNAs into HEK-293T and collected sEVs from cell culture conditioned medium. We performed RT-qPCR to test whether siRNA was loaded into sEVs.
Plasmids expressing β-catenin mRNA and L7Ae-CD63 were co-transfected into HEK-293T and cell culture medium was collected to extract sEVs. sEV’s RNA was extracted to detect whether the mRNA was embedded into them by RT-qPCR.
At the same time, we purified sEVs by three different methods (UC, EIK, and TFF/BE-SEC). To assess the purity of sEVs, we performed NanoSight to detect the size distribution of sEVs and calculated the number of particles per microgram of protein. We also conducted WB to assess the abundance of EV surface markers (CD9, CD63, TSG101). After comprehensive comparation of these three methods, we determined TFF/BE-SEC has the best EV purity.
Finally, we validated that our therapeutic sEVs (which carry 5αR-siRNA-1, Piezo1-siRNA-5 and β-catenin mRNA) purified by TFF/BE-SEC could inhibit the apoptosis of DPC induced by androgen using flow cytometry.
On the NCBI web page, we designed the required murine siRNA for 5αR and Piezo1. The followings are our results:
siRNAs for 5αR:
5αR-siRNA-1: AAUGAGUAAAUAAAUGUCCUG
5αR-siRNA-2: AAUAAACCAGGUAAUAGGCUU
5αR-siRNA-3: AACAAAGUGUGAAAAAUGCAA
5αR-siRNA-4: UCAGAAAGAUCACCGCUGAUA
5αR-siRNA-5: UAAACCAGGUAAUAGGCUUGC
5αR-siRNA-6: AAACAAGCCACCUUGUGGGAU
siRNAs for Piezo1:
Piezo1-siRNA-1: UAGAAACAGCAAAUAGACCAG
Piezo1-siRNA-2: AGUAUAGGCAAAUGAGAUGGC
Piezo1-siRNA-3: AUAAAUGGUGUCUGAUAGCAG
Piezo1-siRNA-4: UAUGUCUUCAUCGUCGUCAUC
Piezo1-siRNA-5: UUCAUCGUCGUCAUCAUCGUC
Piezo1-siRNA-6: AUCAUCGUCAUCGUCAUCAUC
We respectively transfected plasmids of 5αR-siRNA into RM-1. Then we conducted RT-qPCR utilizing total cellular RNA and performed WB using total proteins. Finally, we found that siRNA-5αR-1 was effective and could reduce the expression of 5αR mRNA by nearly 80 percent. This result indicates that our siRNA can successfully knock down the expression of 5αR.
We respectively transfected plasmids of Piezo1-siRNA into RM-1. Then we conducted RT-qPCR utilizing total cellular RNA and performed WB using total proteins. Finally, we found that Piezo1-siRNA-5 was the most effective and could reduce the expression of Pizeo1 mRNA by nearly 70 percent. This result indicated that our siRNA can successfully knock down the expression of Piezo1.
After screening out the effective siRNAs (5αR-siRNA-1 and Piezo1-siRNA-5), we transfected the corresponding plasmids into HEK-293T and purified sEVs from cell culture medium. Then we conducted RT-qPCR utilizing total RNAs of the specific sEVs and proved that the sEVs do have the siRNAs we need.
After reading the literature [2,3], we found the required nucleotide sequences on NCBI and inserted them into pcDNA3.1-mCherry to construct the plasmids (pcDNA3.1-box CD mini-β catenin-mCherry and pcDNA3.1-L7Ae-G8Linker-CD63-mCherry).
The plasmid pcDNA3.1-box CD mini-β catenin-mCherry was transfected into HEK-293T, and the total cellular RNAs were used for RT-qPCR to verify the expression level of β-catenin mRNA. The result indicated that our plasmid could express β-catenin.
We co-transfected the two plasmids (pcDNA3.1-box CD mini-β catenin-mCherry and pcDNA3.1-CD63-L7Ae-mCherry) into HEK-293T and collected sEVs from cell culture medium. The total RNA of the specific sEVs was extracted for RT-qPCR to detect mRNA-β-catenin. The result demonstrated that our approach of wrapping mRNA-β-catenin into sEVs is feasible.
We comprehensively evaluated the EV yields and sample purities of three most popular EV separation methods, ultracentrifugation, commercial sEV isolation kit, and tangential flow filtration combined with bind-elute size exclusion chromatography (BE-SEC) in cell culture medium.
Namely, we performed particle size analysis of sEVs using NanoSight. To assess the purity of the sEVs, we counted the number of particles per microgram of protein. The most common impurities in sEVs extraction are soluble proteins outside the vesicles, so the number of particles per microgram of protein can reflect the purity of sEVs. The results showed that the particle size of sEVs extracted by TFF/BE-SEC is mostly concentrated within 50-200 nm, with the highest number of particles per microgram protein and the highest surface marker abundance, indicating the highest purity. Meanwhile, TFF/BE-SEC can extract sEVs from at least 200 ml cell culture media each time, while UC and EIK take longer or even require multiple experiments to process the same amount of culture media. In summary, TFF/BE-SEC is the best method for the large-scale purification of sEVs.
We noted that the flow through collected from BE-SEC column could be divided into three parts. To further investigate which part contains sEVs with the highest purity, we collected them respectively for downstream analysis. As indicated in Fig.10, sEVs from Period 2 are the purest.
We expected to prove that our therapeutic sEVs (which load with 5αR-siRNA-1, Piezo1-siRNA-5, and mRNA-β-catenin) can inhibit apoptosis of DPC in the presence of androgen. Thus, we added testosterone propionate (TP) to the cell culture medium and co-cultured therapeutic sEVs with DPC for 30h. Then the apoptotic status of DPC was detected using Annexin V-FITC Apoptosis Detection Kit (from Beyoncé) by flow cytometry. As shown in Fig.11, TP can induce apoptosis, but apoptosis was significantly inhibited after the addition of sEVs. These findings indicated that our sEVs can be used as a treatment option and may have therapeutic implications for patients with AGA.
Reference:
[1] Dhariwala, M. Y., & Ravikumar, P. (2019). An overview of herbal alternatives in androgenetic alopecia. Journal of cosmetic dermatology, 18(4), 966–975. https://doi.org/10.1111/jocd.12930
[2] Xie Y, Chen D, Jiang K, Song L, Qian N, Du Y, Yang Y, Wang F, Chen T. Hair shaft miniaturization causes stem cell depletion through mechanosensory signals mediated by a Piezo1-calcium-TNF-α axis. Cell Stem Cell. 2022 Jan 6;29(1):70-85.e6. doi: 10.1016/j.stem.2021.09.009. Epub 2021 Oct 7. PMID: 34624205.
[3] Premanand A, Reena Rajkumari B. Androgen modulation of Wnt/β-catenin signaling in androgenetic alopecia. Arch Dermatol Res. 2018 Jul;310(5):391-399. doi: 10.1007/s00403-018-1826-8. Epub 2018 Mar 16. PMID: 29549490.