Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Feb 1.
Published in final edited form as: DNA Repair (Amst). 2021 Dec 30;110:103271. doi: 10.1016/j.dnarep.2021.103271

The mRNA Tether Model for Activation-Induced Deaminase and its Relevance for Ig Somatic Hypermutation and Class Switch Recombination

Di Liu 1,#, Myron F Goodman 2, Phuong Pham 2, Kefei Yu 3, Chih-Lin Hsieh 4, Michael R Lieber 1,*
PMCID: PMC8816865  NIHMSID: NIHMS1768572  PMID: 34990960

Abstract

Activation-induced deaminase (AID) only deaminates cytosine within single-stranded DNA. Transcription is known to increase AID deamination on duplex DNA substrates during transcription. Using a purified T7 RNA polymerase transcription system, we recently found that AID deamination of a duplex DNA substrate is reduced if RNase A is added during transcription. This finding prompted us to consider that the mRNA tail may contribute to AID action at the nearby transcribed strand (TS) or non-transcribed strand (NTS) of DNA, which are transiently single-stranded in the wake of RNA polymerase movement. Here, we used a purified system to test whether a single-stranded oligonucleotide (oligo) consisting of RNA in the 5’ portion and DNA in the 3’ portion (i.e., 5’RNA-DNA3’, also termed an RNA-DNA fusion substrate) could be deaminated equally efficiently as the same sequence when it is entirely DNA. We found that AID acts on the RNA-DNA fusion substrate and the DNA-only substrate with similar efficiency. Based on this finding and our recent observation on the importance of the mRNA tail, we propose a model in which the proximity and length of the mRNA tail provide a critical site for AID loading to permit a high local collision frequency with the NTS and TS in the transient wake of the RNA polymerase. When the mRNA tail is not present, we know that AID action drops to levels equivalent to when there is no transcription at all. This mRNA tether model explains several local and global features of Ig somatic hypermutation and Ig class switch recombination, while integrating structural and functional features of AID.

Keywords: activation-induced deaminase (AID), class switch recombination (CSR), somatic hypermutation (SHM), R-loop, immunoglobulin, cytosine deaminase

1. Introduction

Activation-induced deaminase (AID) is the essential cytosine deaminase required to initiate a series of steps for both immunoglobulin (Ig) class switch recombination (CSR) and somatic hypermutation (SHM) [1]. AID only deaminates C within single-stranded DNA, and therefore transcription is needed to achieve efficient AID action on duplex DNA [2, 3]. Ig CSR occurs only at the Ig heavy chain locus (IgH) at repetitive class switch regions containing a high density of G-clusters on the non-transcribed strand (NTS) and AID preferred sites consisting of a C within the configuration WRC or WRCW (where W = A or T and R = A or G). AID can act at sub-optimal C’s with up to 10-fold reduced efficiency within DNA, but cannot act on RNA [2, 3]. While Ig SHM can occur at any transcribed DNA sequence, it occurs with primary physiological importance at the IgH and Ig light chain (IgL) variable domain exons. Action of AID at off-target locations is well-described and is heavily reliant on transcription [4, 5].

Under optimal conditions, each AID molecule binds two strands of single-stranded nucleic acid, one in the assistant patch and one in the substrate (catalytic) channel, in order to act on the DNA in the substrate channel [6]. In the absence of transcription, AID is able to act on single-stranded DNA of 24 nt or longer in length [7]. Many laboratories have proposed that upon transcription, the increased transient opening of the DNA duplex can lead to higher AID activity on the NTS; and the slowdown of the RNA polymerase may also increase the chance of two nearby ssDNA regions being captured by a single AID molecule and lead to an increase in AID activity.

Recently, we found that addition of RNase A to a defined T7 transcription system containing purified AID had a strong effect on AID deamination of the transcribing duplex [8]. Based on this new finding, we consider an additional possible mechanism: multiple AID may bind to a single RNA transcript, and these AID molecules could then act on the local NTS or TS DNA. In this model, AID loads onto the mRNA tail via the assistant patch, thus enabling the AID substrate patch to act on the ssDNA that arises on the NTS and TS in the wake of the RNA polymerase movement. This model is based on the current understanding of the following: a) AID can bind to RNA (but not deaminate within RNA) [6]; b) AID can bind to ssDNA [6]; c) both assistant patch and substrate channel have to be occupied for AID to deaminate a DNA molecule harboring an AID hotspot [6]; d) RNase A presence during transcription decreases AID deamination [8]; e) both the assistant patch and the substrate channel of AID may bind to RNA even though AID does not act on RNA [6]. Here we test this model using a defined system, and we discuss the broader implications for an mRNA tether model in the AID mechanism during Ig CSR and SHM.

2. Methods and materials

2.1. Oligonucleotides and DNA substrates

DNA oligonucleotides (oligos) were synthesized by Integrated DNA Technologies, purified by HPLC method, and directly used in the reactions. The 30 nt DL133 RNA-DNA fusion oligo (5’-UUUUUUUUUUUUUTTTTAACTTTTTTTTTT-3’), 30 nt DL134 DNA oligo (5’-TTTTTTTTTTTTTTTTTAACTTTTTTTTTT-3’), and 17 nt DL135 DNA oligo (5’-TTTTAACTTTTTTTTTT-3’) contained a FAM label at 3’ end. The 30 nt DL136 (5’-rArArArArArArArArArArArArAAAAAAACAAAAAAAAAA-3’) was a 5’RNA-DNA3’ fusion substrate with 13 nt phosphorothioated ribonucleotides and 17 nt deoxynucleotides containing a 5’ IRD700 label. A single cytosine was at position 20, 20, 7, and 20 of DL133, DL134, DL135, and DL136, respectively.

2.2. AID deamination reactions

GST-AID was purified as previously described [9]. Briefly, Sf9 insect cell expressed GST-AID was first purified with Ni-NTA resin and then pre-activated by RNase A. RNase A was washed away by extensive washes after binding GST-AID to Glutathione Sepharose resin. In a typical deamination reaction, the ratio of substrate to GST-AID was kept at 0.25:1. The 10 μl reactions containing 100 fmol/μl of GST-AID and 25 fmol/μl of oligo substrate were incubated in the reaction buffer (10 mM Tris-HCl, 0.1 mM EDTA, pH 8.0) at 37°C for 10 min. RNase inhibitor (New England Biolabs) was added when indicated to a final concentration of 1 unit/μl. Deamination reactions were supplemented with 0.5 unit of USER enzyme (New England Biolabs) and incubated at 37°C for 15 min. Equal volume of formamide were then added and samples were heated at 100°C for 5 min and plunged on ice for 5 min.

The deamination reactions with various substrate to GST-AID ratios (0.25:1, 1:1, and 4:1) were prepared with constant AID concentration at 100 fmol/μl and various substrate oligo amounts (substrate masses of 0.25, 1 and 4 pmol; substrate concentrations of 0.125, 0.5 and 2 uM). The reactions were performed in the same manner as mentioned above and a portion from each reaction mix containing 250 fmol of the oligo substrate (DL134, DL135, and DL136) was loaded onto the denaturing PAGE with size ladders. The deamination percentage was calculated by dividing the intensity of the deamination product band by that of the whole lane.

The deamination reactions shown in the supplementary results were not treated with USER enzyme following AID treatment but were incubated with 1 unit of uracil DNA glycosylase (UDG, Invitrogen) at 37°C for 30 min followed by addition of 1 μl of 2 M NaOH and heated for 5 min at 95°C.

2.3. Quantitation of AID Deamination Activity

The AID deamination reaction products were separated on 20% denaturing polyacrylamide gels (PAGE). Gels were visualized by iBright FL1000 Gel (ThermoFisher) in fluorescent blots mode using two channels with 488 nm and 700 nm excitation wavelengths. The percentage of the deaminated product was calculated by dividing the quantitation of fluorescence in the deaminated product band by the total quantitation of fluorescence in the lane after background subtraction using the ImageJ software (NIH; [10]).

The expected percentages of deamination product in the reactions of substrate to AID ratios at 1:1 and 1:4 were calculated based on the relative rate of deamination in the lowest substrate to AID ratio (0.25:1) for a 1st-order reaction and a 2nd-order reaction. The relative rate of deamination in the reaction was calculated by multiplying the percentage of deaminated product by the total number of substrate molecules in the reaction. For the expected percentage of deamination if the reaction is of 1st-order when the substrate:AID ratio is 1:1, the rate of deamination in the substrate:AID ratio at 0.25:1 was multiplied by the 4X increase of substrate concentration and then divided by the total concentration of substrate in the starting reaction. For the expected percentage of deamination if the reaction is of 1st-order when substrate:AID ratio is 4:1, the rate of deamination in the substrate:AID ratio at 0.25:1 was multiplied by the 16X increase of substrate concentration and then divided by the total concentration of substrate in the starting reaction. For the expected percentage of deamination if the reaction rate is 2nd-order, a 16X factor (square of 4X increase in substrate concentration) and a 256X factor (square of 16X increase in substrate concentration) were used when the substrate:AID ratios were 1:1 and 1:4, respectively.

3. Results

3.1. AID can deaminate deoxycytidine on both DNA and the DNA portion of RNA-DNA fusion substrates

Since each AID molecule has both an assistant patch and a substrate channel for single-stranded nucleotide acid binding [6], we wondered if an RNA molecule can be a partner of DNA by threading through the assistant patch, while DNA is deaminated in the substrate (i.e., catalytic) channel. Deamination reactions with DNA only or with RNA-DNA fusion substrates, both containing a single cytosine at position 20 from the 5’ end, were designed according to the orientation of the nucleic acid in the assistant patch and the substrate channel demonstrated in the atomic structure [6]. In the 5’RNA-DNA3’ fusion substrate, the RNA and DNA are on the same oligo strand with 13 nt of RNA at the 5’ end and 17 nt of DNA at the 3’ end.

A 0.25:1 substrate to AID ratio was used in the reactions (Fig. 1). We see that AID deamination of the single cytosine on the 3’ FAM labeled DNA-only substrate resulted in a 10 nt band (Fig. 1, 10 nt position indicated by the white arrow). This is the expected position of the deamination product, given the 3’ FAM label. A 20 nt AID deamination product was observed in the reactions using the 5’ labeled RNA-DNA fusion substrate. Given the 5’ IRD label, this is the expected position for AID action at the AID hotspot. As controls, AID reactions using a 17 nt DNA substrate with a single cytosine at position 7 did not show any deamination products (Fig. S1A), suggesting that the DNA portion alone in the RNA-DNA fusion substrate is insufficient for AID deamination.

Figure 1. AID can deaminate cytosines on both a DNA substrate and an RNA-DNA fusion substrate.

Figure 1.

Open in a new tab

(A) AID deamination results of 3’ FAM labeled DNA substrate and 5’ IRD700 labeled RNA-DNA hybrid substrate. The deamination reactions were performed with a 0.25:1 of substrate to AID ratio using a 3’ FAM labeled DNA substrate (reactions #1 to #4, in green) and a 5’ IRD700 labeled RNA-DNA fusion substrate (reactions #5 to #8, in red) in presence and absence of RNase inhibitor. AID deamination of the DNA substrate results in a 10 nt band with FAM label, while a 20 nt band with IRD700 label is generated from AID deamination of the RNA-DNA fusion substrate, as indicated by the white arrows. Equal amount of DL134, DL135, and DL136 as in the deamination reactions were used as size markers. (B) Separate images of FAM and IRD700 channels for the deamination reactions. Figure 1 was done three times with equivalent results, and the figure shown is representative. In addition, two additional experiments were very similar in design and these also were fully consistent with the gels shown.

The AID used in this assay was pretreated with RNase A to remove insect cell RNA bound to AID during purification, and this was followed by thorough washing to remove RNase A [2, 3, 9]. It is inferred that these RNA molecules are bound to both the assistant patch and substrate channel of AID, since AID is inactive without the removal of these molecules. However, there is still trace residual RNA in the AID as well as trace residual RNase A in the purified AID protein preparation despite the RNase A treatment and extensive washing to remove RNase A [3] (Fig. S1B). There is some indication that addition of RNase inhibitor reduces the AID deamination of DNA oligos [3, 9]. When RNase inhibitor is added to the specified deamination reactions to inhibit the residual RNase A in the purified AID protein preparation (Fig. 1, lane 4), the 10 nt band derived from deamination is less intense than what is observed in the reaction without RNase inhibitor (Fig. 1, lane 3). It is possible that the residual RNase A in the purified AID can increase AID deamination by continuing to remove the residual RNA released from the AID and making the assistant and substrate channels available to the DNA substrate (so that the residual RNA does not compete with the DNA substrate). However, we did not observe decreased deamination activity upon addition of RNase inhibitor for the RNA-DNA fusion substrate. This lack of difference may be caused by the loss of IRD700 label at the 5’ end due to RNase A cleavage of the RNA portion of the substrate in reactions without RNase inhibitor (Fig. 1, lane 7).

These results indicate that the assistant patch of AID can bind to either RNA or DNA and allow AID to deaminate the preferred DNA site in the substrate channel. Furthermore, AID does not show any apparent difference in ability to deaminate C within the DNA-only substrate versus C within the DNA portion of the RNA-DNA fusion substrate.

3.2. Stoichiometry of AID binding in the deamination reaction

We hypothesized that one single-stranded DNA oligo of sufficient length (i.e., 30 nt) can occupy both the assistant patch and the substrate channel at the same time [7], in which case, the rate of the deaminated product formation would be 1st-order with respect to the substrate concentration. This is because one portion of the oligo would initially bind first to either the assistant patch or the substrate channel. Once bound, the other portion of the oligo would bind with zero-order kinetics to the remaining patch or channel. However, it is possible that two independent oligos can occupy the two binding sites of a single AID molecule, in which case the reaction would be 2nd-order with respect to substrate concentration.

To test the hypothesis that AID interaction with a 30 nt oligo has a 1st-order rate, reactions with increasing concentrations of RNA-DNA fusion substrate and a constant concentration of AID were performed over a 16-fold range of AID to substrate ratio (Fig. 2). Where specified, RNase inhibitor was used to inhibit the activity of residual RNase A in the purified AID protein to prevent the potential loss of IRD700 label at the 5’ end of the RNA-DNA fusion substrate. To ensure that the amount of the oligo on the gel would remain within the linear range of the fluorescent signal, equal amounts of 250 fmol total oligo from each of the reactions were loaded on the denaturing gel.

Figure 2. AID deamination results at various substrate to AID ratios.

Figure 2.

Open in a new tab

AID deamination reactions with different substrate to AID ratios using the 5’ IRD700 labeled RNA-DNA fusion substrate. The deamination reactions contained constant AID concentration at 100 fmol/μl and various substrate concentrations. Three RNA-DNA fusion substrate to AID ratios, 0.25:1, 1:1, and 4:1, were tested with (#5 to #8) and without (#1 to #4) RNase inhibitor. The portion of deamination reactions containing 250 fmol substrate was loaded from each reaction onto the gel. A 20 nt deamination product as indicated by the red arrow was expected from AID deamination of the 5’ labeled fusion substrate.

We find that the amount of deamination product as a percentage of total oligo is 4.3%, 10.6%, and 7.1% for the three reactions with RNase inhibitor that have a substrate to AID ratio of 0.25:1, 1:1, and 4:1, respectively (Fig. 3A, Exp1). The percentage of deamination product is 2.9%, 7.1%, and 7.6% for the three reactions without RNase inhibitor that have a substrate to AID ratio of 0.25:1, 1:1, and 4:1, respectively (Fig. 3B, Exp1). Repeated independent experiments show similar results (Fig. 3, Exp2). To evaluate whether these findings support the 1st-order reaction, we compare the experimental results with the expected percentage of deamination for a 1st-order versus a 2nd-order reaction (Fig. 3). The expected percentage of deamination product for a 1st-order reaction and a 2nd-order reaction was calculated using the rate of deamination from the reaction of 0.25:1 substrate to AID ratio in the respective experiments as described in the Materials and Methods. For example, the rate of deamination in the 1:1 reaction and the 1:4 reaction would be 4X and 16X higher than in the 0.25:1 reaction, respectively, but the percentage of deamination would be the same for all three reactions if the reaction is of 1st-order. If the reaction is of 2nd-order, the rate of deamination in the 1:1 reaction and the 1:4 reaction would be 16X and 256X higher than in the 0.25:1 reaction, respectively, and the percentage of deamination would be 4X and 16X higher in the 1:1 reaction and in the 1:4 reaction than in the 0.25:1 reaction, respectively. Despite minor experimental variations, it is clear that these reactions are not of 2nd-order with respect to substrate concentration. The reactions are consistent with a 1st-order dependence on substrate concentration, and strongly support the hypothesis that the two patches on AID are occupied by the same 30 nt oligo substrate molecule, regardless of whether this is DNA-only or an RNA-DNA fusion substrate.

Figure 3. Quantitation of AID deamination product demonstrates that the reaction is a 1st-order reaction with respect to substrate concentration.

Figure 3.

Open in a new tab

(A) Experiments with the addition of RNase inhibitor. (B) Experiments with no RNase inhibitor added. Percentage of deamination product from reactions with different substrate to AID ratios were quantitated using the ImageJ software and summarized above the plot of each set of experiments. The expected percentage of deamination product for each of the reactions as a 1st-order and a 2nd-order reaction (with substrate to AID ratios of 1:1 and 1:4) was calculated based on the percentage of deamination product for substrate to AID ratio of 0.25:1 from the respective experiments as described and also listed above the plot of each set of experiments. In the plots, blue color and red color represent the experimental (solid line with crosses) and the expected percentages of deamination product for 1st-order (dashed line with triangles) and 2nd-order reactions (dotted line with solid circles) for two different experiments, Exp1 and Exp2, respectively.

4. Discussion

4.1. Summary and Model

We recently observed that AID deamination of a transcribing duplex DNA target in a purified system decreased markedly when RNase A was added to remove the mRNA tail from the RNA polymerase [8]. As we considered the basis for this, the most likely explanation was that the mRNA tail plays a key role in locally directing AID to its DNA target. Here, we have tested whether proximally positioned RNA could support AID action at a nearby DNA target (i.e., in the 5’RNA-DNA3’ fusion substrate). We found that AID deaminations within a 5’RNA-DNA3’ substrate occurred as efficiently as deamination within the same length target consisting entirely of ssDNA. In addition, we also found that the AID deamination activity is most likely a 1st-order reaction with respect to substrate concentration.

If RNA can be equally effective for binding to the assistant patch of AID for productive deamination of DNA, what role might the RNA have for AID deamination in the cells where DNA is primarily duplex? Each molecule of AID requires its assistant patch binding to single-stranded nucleic acid [6], and its substrate channel can act on any ssDNA containing a C, optimally within a WRC motif. As known, a DNA region with a WRC motif must be single-stranded to be deaminated by AID [2, 3]. In the human genome, the RNA polymerase generates a transient negative superhelical wave in the wake of its movement, and this leaves the NTS and TS transiently exposed in single-stranded form for AID action [11, 12]. The mRNA tail from the transcription is single-stranded and is readily available to bind to the assistant patch of AID, and this is precisely where transient single-stranded DNA is being generated by the RNA polymerase. In contrast, nearly all the other nearby DNA is duplex and unavailable for AID deamination. Therefore, the binding of AID to the mRNA has distinct advantages for AID deamination of the upstream single-stranded DNA. Moreover, several AID molecules could load onto the long mRNA as it exits the RNA polymerase, thus improving the collision frequency of any one of these AID molecules with any C within the NTS or TS in the transient ssDNA wake.

4.2. Relevance of the Model to AID Action During Ig SHM

Our model would explain an important aspect of SHM that has remained unexplained for several decades. SHM does not start immediately downstream of the transcription start site. Oddly, it starts ~100 bp downstream of the IgH and IgL chain gene promoters [13]. This could be explained by two unique aspects of our model: the length of the mRNA tail for AID binding and the transient negative superhelical tension due to transcription. AID has a hydrodynamic radius of ~32 angstroms, which means that 100 nt long mRNA could bind many AID monomers. Transient negative superhelical tension extends upstream [11, 12, 14]. A length of 100 bp is ~ 10 turns of the helix [11], and the greatest transient negative superhelical tension is within a similar length [14].

AID action at non-Ig locations is similar to AID action during SHM at Ig loci in that it correlates with transcription [4, 5]. In purified biochemical systems, AID action at non-Ig sequences is also reliant on the mRNA tail, and removal of the mRNA tail results in a marked decrease in deaminations down to background levels [8].

4.3. Relevance of the Model to AID Action During Ig CSR

The relevance of our model to CSR is also clear because an R-loop forms at each switch region, and the NTS remains in a ssDNA state for a prolonged period [15, 16]. Furthermore, upon action of RNase H, the R-loop can collapse to generate heteroduplex regions on the NTS and TS when repeats mis-align. AID bound to the mRNA could act at the ssDNA in these heterologous loops [15].

4.4. Relation to Other Models for AID Action

Elegant work has been summarized previously for association of transcription and AID action [17], and the models have invoked binding of AID to the RNA polymerase. Binding of AID to G4-RNA also has been proposed, followed by transfer of the AID protein to nearby G4-DNA [18, 19]. Secondary structures formed by the DNA or RNA, with subsequent events needed for AID action, have also been proposed [6].

Lack of binding of AID directly to RNA polymerases derived from prokaryotes or bacterial viruses argues against such a requirement for AID binding directly to the RNA polymerase to achieve cytosine deamination [20]. Regarding nucleic acid secondary structures, the action of AID during SHM is inconsistent with these models because there are no G4-forming sequences for RNA or DNA consistently found at the location at which SHM occurs, and all models for G4-DNA or G4-RNA secondary structure acknowledged that the models failed to explain AID targeting in SHM [19].

In contrast, our much simpler model proposes AID action during SHM, CSR, and non-Ig locations in a similar manner. The model formulation is driven by the AID crystal structures [6] and our simple observation of removing the nascent RNA tail during transcription [8]. Of note, earlier studies in which RNA polymerase movement paused, but the RNA tail was not removed, also increased AID deamination rate. This is consistent with increased frequency of AID collision with the DNA target [21], and therefore, very consistent with our model, though binding to the mRNA tail was not proposed at that time.

One previous study performed in vitro AID biochemical activity and binding assays comparing oligo substrates containing a DNA-RNA hybrid bubble versus an equivalent DNA-DNA bubble. They found that some mutations in the assistant path were more likely to affect binding to the DNA:DNA oligo than DNA:RNA oligo [22]. Their suggestion that RNA could increase the activity of AID at DNA:RNA hybrid regions during transcription would support our mRNA tethering model.

4.5. Experimental Complexities for Testing the Model and Limitations of Our Study

In earlier work on recombinant AID, addition of random RNA decreased AID activity on ssDNA [2, 3]. It is interesting to consider the molar ratio of the RNA and the ssDNA substrate. The RNA-free AID was incubated with a variety of long RNAs at molar ratios as high as 1:20 (AID:RNA), and at these high substrate ratios, the AID may be bound to two RNA molecules and be fully inhibited from binding to ssDNA. The RNA transcript also may inhibit AID if the RNA molecule is long enough to thread through both binding sites of a single AID molecule. An additional possibility is that RNA may have higher affinity for AID and can out- compete DNA for binding, if all interactions occur in trans. However, in our current study, RNA does not appear to bind to AID any more efficiently; otherwise, it would show more deamination in the RNA-DNA fusion substrate than the DNA substrate. Our results support the model here that RNA binding facilitates AID deamination of the nearby DNA. Nevertheless, additional detailed study by many different physical techniques and by co-crystallization of AID with RNA and DNA of various lengths will be needed to fully evaluate the validity of our mRNA tether model for AID action.

Given the orientation requirements of the assistant patch and the catalytic channel, we think it is likely that the RNA portion of our RNA-DNA fusion oligo occupies the assistant patch. Further work in the AID field will be needed to provide direct evidence of such positioning.

Others reported that certain AID variants with mutations in the assistant patch did not affect catalytic activity on a bubble DNA oligo substrate [23]. These variants can mutate the E. coli rpoB gene with equal efficiency as the wild type. Since E. coli mutation by AID is transcription-dependent [24], one could wonder why the assistant patch would be dispensable for transcription-linked deamination in E. coli. Relevant to this, the Goodman lab has described different patterns of AID deamination on T7 transcribed templates versus mammalian RNA pol II-mediated transcription [9].

Regarding cellular tests of our model, given the abundance of RNA in the nucleus and cytoplasm in eukaryotes, it is difficult to purify AID from any cellular source without any RNA at all. Similarly, it is difficult to design cellular experiments that test for AID binding to the mRNA tail. Expression of RNase enzymes in cells might reduce AID action, though toxicity is likely to be an issue. For these reasons, we felt that it was important to propose the model for AID action described here to encourage testing, since it may require many lines of indirect evidence to fully test it in the future.

Supplementary Material

1

Figure 4. Proposed mechanisms of AID activity during transcription.

Figure 4.

Open in a new tab

(A) AID deamination of the Y-shape substrate. AID is shown as a green solid circle, and the Y-shape substrate is shown in blue. The two ligand binding channels, assistant patch and substrate channel, of AID are shown in yellow and red correspondingly. The catalytic site within the substrate channel is shown as a solid black dot. The substrate lies in a 5’ to 3’ direction from the branch point of the Y-shape structure to the catalytic site of AID in the substrate channel. (B) AID deamination mechanism on the non-transcribed strand during transcription. The non-transcribed strand (NTS) and transcribed strand (TS) of the DNA are shown as blue lines and the nascent RNA transcript is shown as grey line. The assistant and substrate channels of AID are shown in yellow and red correspondingly. AID binds to the RNA transcript through the assistant patch (the yellow line) and to the NTS through the substrate channel (the red line). Deamination of cytosines on the NTS by AID could occur during transcription with the assistant of RNA transcript.

Highlights.

  1. AID deaminates DNA in a fusion substrate with RNA in the 5’ and DNA in the 3’ end.

  2. AID action on DNA-only versus RNA-DNA substrates of equal length are the same.

  3. We propose a model where the mRNA tail recruits AID during transcription.

  4. This model would explain key aspects of Ig somatic hypermutation and class switch.

Acknowledgements:

The authors thank all members of their respective labs for helpful discussions.

Funding and additional information:

This work was supported by NIH grants to MRL (CA100504, GM118009) and MFG (5R35ES028343, GM130450).

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Conflict of interest: The authors declare that they have no conflicts of interest with the contents of this article.

Data availability statement:

All data generated and analyzed during this study are included in this article and its supporting information.

References

  • [1].Muramatsu M, Sankaranand V, Anant S, Sugai M, Kinoshita K, Davidson NO, Honjo T, Specific expression of activation-induced cytidine deaminase (AID), a novel member of the RNA-editing deaminase family in germinal center B cells, J Biol Chem, 274 (1999) 18470–18476. [DOI] [PubMed] [Google Scholar]
  • [2].Pham P, Bransteitter R, Petruska J, Goodman MF, Processive AID-catalysed cytosine deamination on single-stranded DNA simulates somatic hypermutation, Nature, 424 (2003) 103–107. [DOI] [PubMed] [Google Scholar]
  • [3].Bransteitter R, Pham P, Scharff MD, Goodman MF, Activation-induced cytidine deaminase deaminates deoxycytidine on single-stranded DNA but requires the action of RNase, Proc Natl Acad Sci U S A, 100 (2003) 4102–4107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [4].Liu M, Duke JL, Richter DJ, Vinuesa CG, Goodnow CC, Kleinstein SH, Schatz DG, Two levels of protection for the B cell genome during somatic hypermutation, Nature, 451 (2008) 841–845. [DOI] [PubMed] [Google Scholar]
  • [5].Storb U, Shen HM, Nicolae D, Somatic hypermutation: processivity of the cytosine deaminase AID and error-free repair of the resulting uracils, Cell Cycle, 8 (2009) 3097–3101. [DOI] [PubMed] [Google Scholar]
  • [6].Qiao Q, Wang L, Meng FL, Hwang JK, Alt FW, Wu H, AID Recognizes Structured DNA for Class Switch Recombination, Mol Cell, 67 (2017) 361–373 e364. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [7].Yu K, Huang FT, Lieber MR, DNA substrate length and surrounding sequence affect the activation-induced deaminase activity at cytidine, J Biol Chem, 279 (2004) 6496–6500. [DOI] [PubMed] [Google Scholar]
  • [8].Liu D, Loh Y-WE, Hsieh CL, Lieber MR, Mechanistic basis for chromosomal translocations at the E2A gene and its broader relevance to human B cell malignancies, Cell Reports, 36 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [9].Pham P, Malik S, Mak C, Calabrese PC, Roeder RG, Goodman MF, AID–RNA polymerase II transcription-dependent deamination of IgV DNA, Nucleic Acids Research, 47 (2019) 10815–10829. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [10].Rasband WS, ImageJ, https://imagej.nih.gov/ij/, in, U. S. National Institutes of Health, Bethesda, Maryland, USA, 1997–2018. [Google Scholar]
  • [11].Sinden RR, DNA structure and function, Gulf Professional Publishing, 1994. [Google Scholar]
  • [12].Liu LF, Wang JC, Supercoiling of the DNA template during transcription, Proc Natl Acad Sci U S A, 84 (1987) 7024–7027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [13].Lebecque SG, Gearhart PJ, Boundaries of somatic mutation in rearranged immunoglobulin genes: 5’ boundary is near the promoter, and 3’ boundary is approximately 1 kb from V(D)J gene, J Exp Med, 172 (1990) 1717–1727. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [14].Kouzine F, Gupta A, Baranello L, Wojtowicz D, Ben-Aissa K, Liu J, Przytycka TM, Levens D, Transcription-dependent dynamic supercoiling is a short-range genomic force, Nat Struct Mol Biol, 20 (2013) 396–403. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [15].Yu K, Lieber MR, Current insights into the mechanism of mammalian immunoglobulin class switch recombination, Crit Rev Biochem Mol Biol, 54 (2019) 333–351. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [16].Yu K, Chedin F, Hsieh CL, Wilson TE, Lieber MR, R-loops at immunoglobulin class switch regions in the chromosomes of stimulated B cells, Nat Immunol, 4 (2003) 442–451. [DOI] [PubMed] [Google Scholar]
  • [17].Storb U, Why does somatic hypermutation by AID require transcription of its target genes?, Adv Immunol, 122 (2014) 253–277. [DOI] [PubMed] [Google Scholar]
  • [18].Pham P, Afif SA, Shimoda M, Maeda K, Sakaguchi N, Pedersen LC, Goodman MF, Structural analysis of the activation-induced deoxycytidine deaminase required in immunoglobulin diversification, DNA repair, 43 (2016) 48–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [19].Pucella JN, Chaudhuri J, AID Invited to the G4 Summit, Mol Cell, 67 (2017) 355–357. [DOI] [PubMed] [Google Scholar]
  • [20].Senavirathne G, Bertram JG, Jaszczur M, Chaurasiya KR, Pham P, Mak CH, Goodman MF, Rueda D, Activation-induced deoxycytidine deaminase (AID) co-transcriptional scanning at single-molecule resolution, Nat Commun, 6 (2015) 10209. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [21].Canugovi C, Samaranayake M, Bhagwat AS, Transcriptional pausing and stalling causes multiple clustered mutations by human activation-induced deaminase, The FASEB Journal, 23 (2009) 34–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [22].Abdouni HS, King JJ, Ghorbani A, Fifield H, Berghuis L, Larijani M, DNA/RNA hybrid substrates modulate the catalytic activity of purified AID, Mol Immunol, 93 (2018) 94–106. [DOI] [PubMed] [Google Scholar]
  • [23].Methot SP, Litzler LC, Subramani PG, Eranki AK, Fifield H, Patenaude AM, Gilmore JC, Santiago GE, Bagci H, Côté JF, Larijani M, Verdun RE, Di Noia JM, A licensing step links AID to transcription elongation for mutagenesis in B cells, Nat Commun, 9 (2018) 1248. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [24].Petersen-Mahrt SK, Harris RS, Neuberger MS, AID mutates E coli suggesting a DNA deamination mechanism for antibody diversification, Nature, 418 (2002) 99–103. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1768572-supplement-1.docx (6.5MB, docx)

Data Availability Statement

All data generated and analyzed during this study are included in this article and its supporting information.

RESOURCES