Concatemeric Broccoli reduces mRNA stability and induces aggregates

Abstract

Fluorogenic aptamers are an alternative to established methodology for real-time imaging of RNA transport and dynamics. We developed Broccoli-aptamer concatemers ranging from 4 to 128 substrate-binding site repeats and characterized their behavior fused to an mCherry-coding mRNA in transient transfection, stable expression, and in recombinant cytomegalovirus infection. Concatemerization of substrate-binding sites increased Broccoli fluorescence up to a concatemer length of 16 copies, upon which fluorescence did not increase and mCherry signals declined. This was due to the combined effects of RNA aptamer aggregation and reduced RNA stability. Unfortunately, both cellular and cytomegalovirus genomes were unable to maintain and express high Broccoli concatemer copy numbers, possibly due to recombination events. Interestingly, negative effects of Broccoli concatemers could be partially rescued by introducing linker sequences in between Broccoli repeats warranting further studies. Finally, we show that even though substrate-bound Broccoli is easily photobleached, it can still be utilized in live-cell imaging by adapting a time-lapse imaging protocol.

Fig 1. Broccoli tandem array synthesis and expression.

(A) Cloning schematic of the constructs containing different copy numbers of Broccoli. (B, C, D) 293T cells were transfected with mCherry or 4x, 8x, 16x, 32x, 64x or 128x Broccoli and analyzed 24 hours post-transfection by flow cytometry. Percentage of Broccoli and mCherry positive cells shown in dot plots (B) and mean fluorescence intensity of Broccoli (C) and mCherry (D) compared between constructs. (D) The mean fluorescence intensity (MFI) of Broccoli in the graph was normalized to the 4x Broccoli construct and the MFI of the remaining constructs was calculated as follows: ⦗ (MFI XxBroccoli): (4xBroccoli) ⦘. Data representative of three independent experiments. Statistical analysis performed by t-test (c, d). * P<0.05, ** P<0.01, *** P<0.001.

Fig 2. Broccoli exhibits punctate morphology.

(A) 293T cells were transfected with mCherry, 4x, 8x, 16x, 32x, 64x, 128x Broccoli plasmids and analysed by microscopy for localization and morphology. (Scale bar = 75 μm) (B) Representative higher magnification inserts illustrating subcellular Broccoli distribution (Scale bars = 10 μm).

Fig 3. Broccoli concatamers do not induce p-bodies nor stress granules.

(A) HEK293T cells were transfected with mCherry, 4x, 16x or 128x Broccoli as well as pEGFP-C1 as transfection control. After 24 hours, the cells were fixed and IF stained for DDX6. (B) The area of DDX6 positive spots in HEK293T cells from A was measured. No significant difference between the Broccoli versions was detectable. (3 Replicates; N_mCherry = 3776, N_4xBroccoli = 3552, N_16xBroccoli = 3710, N_128xBroccoli = 5312) (C) HEK293T cells were treated/mock-treated with 0.5 mM Sodium Arsenite for 1h to induce cell stress. Afterwards, the cells were fixed and stained for DDX6. (D) The area of DDX6-positive spots in HEK293T cells from (C) was measured. The difference between the mock treatment and the Arsenite treatment is statistically significant with p = 0.0348. (3 Replicates; N_Arsenite = 9796, N_Mock = 12526) (E) HEK293T cells were transfected and fixed for IF as described in A. The cells were afterwards IF stained for eIF3a. (F) The area of eIF3a positive spots in cells from E was measured. No significant difference between the Broccoli versions was detectable. (4 Replicates; N_mCherry = 6030, N_4xBroccoli = 4674, N_16xBroccoli = 5860, N_128xBroccoli = 6992) (G) HEK293T cells were treated with Sodium Arsenite as described in (C). After fixation the cells were IF stained for eIF3a. (H) The area of eIF3a positive spots in cells from G was measured. The difference between the mock treatment and the Arsenite treatment is statistically significant with p < 0.0001 (3 Replicates; N_Arsenite = 10704, N_Mock = 8200). All error bars show the standard error of the mean and all scale bars represent 20 μm.

Fig 4. RNA of Broccoli is detectable up to 32 copies by Northern blot.

293T cells were transfected with mCherry or Broccoli plasmids and 10μg of isolated total RNA loaded on a denaturing gel for Northern blot. A fluorescently-labelled probe against mCherry (A) or Broccoli (B) was used to detect the respective RNAs. Intensities of mCherry and Broccoli RNA bands were measured using arbitrary values and normalized to either mCherry (A) or 4xBroccoli (B). 16s RNA was stained using ethidium bromide. (C) 293T cells were transfected with mCherry, 32x and 128x Broccoli plasmids. 24hours post-transfection, cells were collected (mock) or treated with actinomycin D for 2 or 6 hours. RNA was isolated at each time point and RT-qPCR performed for detection of mCherry and c-myc. Empty bars show mCherry transcription levels, striped bars represent c-myc transcription levels. Statistical analysis performed by 2-way ANOVA. * P<0.05, ** P<0.01, *** P<0.001.

Fig 5. Polyclonal Broccoli-expressing cell lines showed reduced mCherry expression and no detectable Broccoli fluorescence.

(A) Flow cytometry measurement of mCherry intensity in WT, mCherry or 16xBroccoli 293T or HeLa cells. (B) Northern Blot of mCherry RNA in the stable cell lines. RNA intensity was quantified and calculated relative to the 16S ribosomal band. All quantifications were further compared to the mCherry RNA in the mCherry-293T cells which was normalized to one. (C) RT-qPCR of mCherry. RNA was isolated from the stable cell lines, cDNA prepared and mCherry quantified by qPCR. Delta delta Ct of mCherry was calculated using GAPDH as the reference gene. Afterwards, all values were normalized to mCherry 293T cells. (D) Standard PCR for Broccoli from DNA isolated from 293T and HeLa cells expressing mCherry or 16x Broccoli (four lanes on the left side); lentiviral plasmid used for the generation of the mCherry and 16x Broccoli cell lines (two lanes on the right side). Contrast of gel image was optimized using the substract background function in Fiji (Fiji.sc). Statistical analysis performed by t-test (C). * P<0.05, ** P<0.01.

Fig 6. NIH-3T3 cells infected with Broccoli-MCMV viruses show reduced mCherry expression but no detectable Broccoli fluorescence.

(A, B) NIH-3T3 cells were infected with undetermined MOI of mCherry-MCMV, 16x-MCMV or 32x-MCMV and analyzed by flow cytometry at 2, 4 or 6 hours post-infection. (A) Dot plots show cells at 6 hours post-infection. (B) Histograms and graph show progression of the viral infection/replication as shown by the mean fluorescence intensity of mCherry. (C) Standard PCR for Broccoli from DNA isolated from NIH-3T3 cells infected with mCh-MCMV, 16x-MCMV and 32x-MCMV, at 6 hours post-infection.

Fig 7. Expression profile of Broccoli tandem constructs with exonic inserts.

(A) Cloning schematic of Broccoli with inserted exonic sequences, exon 3 or exon 5, from GAPDH gene of NIH-3T3 cells. (B, C, D) Flow cytometry data of 293T cells transfected with the different plasmids, collected and analysed 24 hours post-transfection. Comparison between the plasmids in regard to percentage of Broccoli and mCherry positive cells (B), mean fluorescence intensity of Broccoli (C) and mean fluorescence intensity of mCherry (D). (C) The values shown were normalized to the cells transfected with 4xBroccoli plasmid. Statistical analysis performed by t-test (C, D). * P<0.05, ** P<0.01.