Biochemical studies of APOBEC protein family

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Request accessible transcript

Transcript (if available)

Content
1

BIOCHEMICAL STUDIES OF APOBEC PROTEIN FAMILY

by

Qihan Chen

A Dissertation Presented to the
FACULTY OF THE USC GRAUDATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY

(Molecular Biology)

May 2016

Copyright 2016 Qihan Chen

2

EPIGRAPH

“Explanations exist: they have existed for all times, for there is always an easy solution to every
problem – neat, plausible and wrong.”
—H.L. Mencken, “The Divine Afflatus”, in the New York Evening Mail, November 16, 1917

3

Dedication

To my family: past, present, and future

4

Acknowledgement
I would like to express my special appreciation and thanks to my advisor Professor Dr. Xiaojiang Chen,
you have been a tremendous mentor for me. I would like to thank you for encouraging my research and
for allowing me to grow as a research scientist. Your advice on both research as well as on my career
have been priceless. I would also like to thank my committee members, professor Lin Chen, professor
Peter Qin for serving as my committee members even at hardship. I also want to thank you for letting
my defense be an enjoyable moment, and for your brilliant comments and suggestions, thanks to you.
I would like to express my thanks to all my former and present lab members: Hanjing Yang, Damian
Wang, Yang Fu , Xiao Xiao, Aaron Wolf, Brett Zirkle, Jiang Gu, Gewen Zhang, Junfeng Wang, Bo
Zhou, Jessica Yu, Ian Slaymaker, Carolyn Truong, Lauren Holden, Ganggang Wang, Maocai Yan,
Kacie Amacher, Hasan Abbas, Meng Xu, Dahai Gai, Braulio Fernandez, Ronda Bransteitter, and
Courtney Prochnow. I am so happy to have you guys around in XJ’s lab. Without their help, I could not
manage to finish.

I would especially like to thank to my dear friends Yaping Liu, Wanqing Yu, Nan Wang, Shi Qiu, Jieli
Sheng, Li Zhou, Rui Wang, Zheda Li, Songze Li, Wei Ma, Xin Song, Wenya Chen, Aina Che, Liang
Jin, Huihui Duan, Xiao Li, Dannuo Chu, Zhuo Zhang, Yunfei Guo, Xiaolu Han, Jiajie Li, Mingyang
Cai, Haipeng Lv, and many more I cannot enumerate here. Your company and support made this
journey truly wonderful.

Finally and forever, I would like to gratefully thank my family. Thank my parents Jianqun Chen and
Yueyu Hang and my girlfriend Yi Shao for their complete and unconditional love! They always try to

5
provide the best and strongest support for my life. Thank my grandparents and aunt for supporting and
caring!

6
Table of contents

EPIGRAPH-----------------------------------------------------------------------------------------2

DEDICATION -------------------------------------------------------------------------------------3

ACKNOWLEDGEMENT-----------------------------------------------------------------------4

TABLE OF CONTENTS ------------------------------------------------------------------------6

LIST OF FIGURES ------------------------------------------------------------------------------8

LIST OF TABLES --------------------------------------------------------------------------------9

ABSTRACT --------------------------------------------------------------------------------------10

CHARPTER 1: INTRODUCTION ----------------------------------------------------------12

1.1 the biochemistry specialty of Double domain APOBECs. ------------------------------12
1.2 the ion effects on APOBECs. ---------------------------------------------------------------13

CHARPTER 2: The in vitro Characterization of APOBEC3F Biochemical
Activities: Importance of Loop 7 on Both CD1 and CD2 for DNA binding and
Deamination --------------------------------------------------------------------------------------15

2.1 RESULTS -------------------------------------------------------------------------------------16
2.2 DISCUSSION ---------------------------------------------------------------------------------27
2.3 METHODS ------------------------------------------------------------------------------------30

7
CHAPTER 3: Divalent Ion Effects on Nine APOBEC Proteins. ----------------------34

3.1 RESULTS -------------------------------------------------------------------------------------35
3.2 DISCUSSION ---------------------------------------------------------------------------------46
3.3 METHODS ------------------------------------------------------------------------------------49

REFERENCE ------------------------------------------------------------------------------------56

8
Figure Index

Figure 2.1. Deamination activities of purified wt and mutant full-length (fl) A3F proteins.--------------17

Figure 2.2. The effects of CD2 loop7 mutations on A3F activities. -----------------------------------------21

Figure 2.3. Roles of CD1 loop 7 W126 for A3F DNA binding and deaminase activities. ---------------26

Figure 2.4 Gel images of ssDNA binding assay of A3F FL wt, CD2 wt and their loop 7 mutants.------33

Figure 3.1. Deamination activities of A3B, A3C, A3F and A3H with and without extra ions.-----------38

Figure S3.1. Deamination activities of AID, Apo1, A3A, A3D, and A3G with and without extra ions. --
-------------------------------------------------------------------------------------------------------------------------39

Figure S3.2. Dose response deamination activity curves of nine APOBEC proteins with and without
extra ions. ------------------------------------------------------------------------------------------------------------40

Figure 3.2. Deamination activity and ssDNA binding of A3F FL wt in the presence of Ca
2+
. ----------41

Figure 3.3. Deamination activity and ssDNA binding of A3C with Ca
2+
and Zn
2+
. -----------------------43

Figure 3.4 ssDNA binding of A3C wt and A3B FL wt with and without extra ions. ---------------------46

Figure S3.3 Coomassie gel images of nine APOBEC proteins and their inactive mutants. --------------52

Figure S3.4 Gel images of EMSA assay of A3F FL wt with various amounts of Ca
2+
. ------------------53

9
Table Index

Table 2.1. Deaminase activity of purified proteins of various A3F constructs. ----------------------------19

Table 2.2. ssDNA (30 nt) binding affinity of purified proteins of various A3F constructs. --------------23

Table 3.1. Comparison of deamination activities of nine APOBEC proteins with and without addition
of divalent ions. -----------------------------------------------------------------------------------------------------36

10

Abstract
APOBEC3F (A3F) is a member of the apolipoprotein B mRNA-editing enzyme catalytic polypeptide-
like (APOBEC) family of proteins that can deaminate cytosine (C) to uracil (U) on nucleic acids. A3F
is one of the four APOBEC members with two Zn-coordinated homologous cytosine deaminase
domains (CDs), with the others being A3B, A3DE and A3G.

Here we report the first comprehensive in vitro characterization of DNA binding and deaminase
activities using purified wild type (wt) and various mutant proteins of A3F from an Escherichia coli
expression system. We show that, even though the N-terminal CD1 domain is catalytically inactive and
the C-terminal CD2 domain is the active deaminase domain, presence of CD1 on the N-terminus of
CD2 enhances the deaminase activity by over an order of magnitude. This enhancement of CD2
catalytic activity is mainly through the increase of substrate ssDNA binding by the N-terminal CD1
domain. We further show that the loop 7 of both CD1 and CD2 of A3F plays an important role for
ssDNA binding for each individual domain as well as for the deaminase activity of CD2 domain in the
full length A3F.

The APOBEC family consists of eleven known members. Nine of them have been shown to deaminate
cytosine (C) to uracil (U) on nucleic acid. The regulation mechanism of APOBEC proteins is still
unclear. Here we demonstrated four divalent ion effects (Ca
2+
, Mg
2+
, Zn
2+
, and Cu
2+
) on nine APOBEC
proteins. With the data of ssDNA binding ability, we revealed that Ca
2+
and Mg
2+
have positive effects
on deamination activity due to the increase of ssDNA binding, while Zn
2+
and Cu
2+
have negative

11
effects on deamination activity due to the decrease of ssDNA binding. In addition, we also showed
Zn
2+
had a special effect on A3C comparing with other APOBEC proteins and ions.

12
Chapter 1: Introduction
The APOBEC family members contain one or two homologous cytosine deaminase (CD) domains that
share a conserved sequence motif (H-X-E-X23-28-P-C-X-C) for Zn-coordination at the putative
catalytic center [1-5]. Four of the eleven family members (A3B, A3DE, A3F, A3G) contain two
homologous CD domains, and the others have only one CD domain. Nine of the eleven members are
shown to catalyze deamination of cytosine to uracil on single-stranded (ss) DNA or RNA [1, 4-6]. Most
of the APOBEC family members constitute parts of the important immune system to safeguard the
potential damage from the intrinsic retroviral elements and infectious retroviral and some other viral
pathogens [7-17]. However, some of these members are also implicated in genome-wide accidental
mutations that may result in significant repercussions and biological consequences such as epigenetic
remodeling, cancer, and even evolution [18-24].

1.1 the biochemistry specialty of Double domain APOBECs.
Among the four APOBEC members containing double CD domains, A3F and A3G show potent anti-
HIV activities [25-29], which is through incorporation/encapsidation of the APOBEC enzymes into
HIV virions [30-33]. The incorporation/encapsidation into HIV virions, which is mainly mediated
through RNA binding by the A3 proteins, requires both CD1 and CD2 domains of A3F, whereas only
CD1 of A3G is needed [31, 33-37]. On the other hand, while HIV Vif targets A3G for E3 ubiquitin
ligase mediated degradation through binding to the CD1 of A3G, Vif targets A3F by binding to its CD2
domain [25, 26, 38-41].

13
The anti-HIV activity is thought to be mostly deamination dependent, with some reports for
deamination independent activity as well [42-47]. Regardless, both CD1 and CD2 domains of A3F and
A3G are required to display anti-HIV activity. For A3F and A3G, CD1 is previously shown to be
catalytically inactive, whereas CD2 is the catalytically active domain for deamination [40, 48-50].
Given the importance of A3F and the knowledge about A3F functions in the literature, there is no
report on the comprehensive in vitro biochemical characterization of the DNA binding and deaminase
activities of A3F using purified A3F proteins. Here we report in vitro biochemical studies of A3F and
various mutant constructs containing loop 7 mutations on either the CD1 or CD2 domains using
purified recombinant proteins from an E. coli expression system.

1.2 the ion effects on APOBECs.
Based on the solved Apobec structures, it is clear that all the members share a highly conserved
structure. Several crystal structures of A3G (3IR2, 3V4J, 3V4K) and A3A (4XXO) show coordination
of a secondary Zinc by loop-3, which is very similar to their common ancestral dCMP deaminase [51,
52]. In addition, two extra Magnesium ions are also shown around in A3G CD2 structure [51]. Both
secondary Zinc and extra Magnesium ions locate at Apobec dimerization interface, which may regulate
its oligomerization status [53]. Previous researches provide little information about how metal ions
affect Apobec proteins in deamination activity and ssDNA binding ability. recent study suggests extra
Zinc ion can increase deamination activities of A3G and A3A by stabilizing the confirmation of
Apobec loop 3. Another study suggested that Mg
2+
may change the bias of A3G deamination process
on the ssDNA from 3’ end to 5’ end as well as Ca
2+
and Mn
2+
[54]. The divalent ion effects were
widely seen in other ssDNA interaction proteins as well, for example, The relaxation reaction activity
of E. Coli. DNA topoisomerase I can be enhanced by Mg
2+
but inhibited by Ca
2+
, Mn
2+
, Co
2+
and Zn
2+

[55]. However, how divalent ions regulate Apobec proteins remains unclear. To address these

14
questions, we characterized the in vitro deamination activities of nine APOBEC proteins (AID, Apo1,
A3A-A3H) with and without common divalent ions Ca
2+
, Mg
2+
, Cu
2+
and Zn
2+
. Our results clarified
that extra Ca
2+
and Mg
2+
enhanced the deamination activities of most APOBEC proteins, while extra
Zn
2+
and Cu
2+
inhibited the deamination activities of most APOBEC proteins. We also shown that
those enhancement and inhibition may due to the change of APOBEC proteins ssDNA binding
abilities. In addition, we demonstrated the special effect of Zn
2+
on A3C deamination activity
comparing with other APOBEC proteins we tested. These study revealed a new direction of APOBEC
proteins regulation by ions, implied another idea of APOBEC-ssDNA co-crystal with ion, and allows
biochemistry studies of APOBEC proteins to be done in a more efficient way with extra ions.

15

Chapter 2: The in vitro Characterization of APOBEC3F Biochemical
Activities: Importance of Loop 7 on Both CD1 and CD2 for DNA
binding and deamination

Reproduced with permission by Qihan Chen, Xiao Xiao, Aaron Wolfe
and Xiaojiang Chen.
Accepted by Journal of Molecular Biology.

Contributions: Q.C. designed the project, analyzed biochemistry functions, and wrote the
manuscript, X.X. helped make mutations and discuss the results, A.W. revised the manuscript and
X.S.C supervised the project.

16
2.1 Results
2.1.1 The role of CD1 and CD2 for A3F deaminase activity.
To assess the in vitro biochemical activities of A3F and the role of loop 7 residues, we first cloned and
purified wild type and mutant A3F proteins to near homogeneity (Figure 2.1A, 2.1B). Four constructs
were tested, including the wild type full length (fl) A3F construct, A3F-fl, as well as the mutants A3F-
fl E67A (with the CD1 catalytic residue mutation at E67), A3F-fl E251A (with the corresponding CD2
catalytic residue mutation), and A3F-fl E67/E251A (with both catalytic residue mutations). These
purified A3F proteins were assayed for deaminase activity using a 30 nucleotide (nt) ssDNA substrate
(Figure 2.1C). As expected, the in vitro activity assay result showed obvious deaminase activity for
both A3F-fl and A3F-fl E67A, whereas no deamination activity was detected for A3F-fl E251A and
A3F-fl E67/E251A (Figure 2.1D, Table 2.1). This in vitro assay result confirms that CD1 has no
deaminase activity, with only CD2 acting as the catalytically active deaminase domain, and is
consistent with previous reports [48]. In addition, quantification of the dose response deamination
activity (Figure 2.1E, 2.1F) revealed that the CD1 mutant A3F-fl E67A activity is more or less
comparable to that of the wild type A3F-fl, suggesting that the E67A mutation within the CD1 domain
did not have a significant impact on the deaminase activity of the full length protein. This is similar to
what is observed for A3B and A3G, where the mutation of the CD1 catalytic Glu residue appears to
have no obvious effect on the fl A3B deaminase activity [48, 56].

17

Figure 2.1. Deamination activities of purified wt and mutant full-length (fl) A3F proteins. (A).
Cartoon representation of the four A3F fl constructs, all expressed as N-terminal MBP fusion proteins.
(B) The SDS-PAGE analysis of the purified A3F proteins from E. coli expression. The proteins are
purified as described in Fu et al.[56]. Briefly, after expressing the A3F proteins in E. coli cells at 16 ℃
for 18 hrs, clarified supernatant of cell lysates in a buffer containing 50 mM Tris-Cl pH 8.0, 150 mM
NaCl, 0.5 mM TCEP was passed through an amylose resin column. After washing the column with 10
column volumes of wash buffer (50 mM Tris-HCL pH 8.0, 1.0 M NaCl, 0.5 mM TCEP, 0.1 μg/ml
RNase A), MBP-A3F proteins were then eluted with buffer containing 20 mM Maltose. The elution
was concentrated and further purified using S200 gel filtration chromatography. The peak fractions
containing the pure MBP-A3F proteins from gel filtration were collected, concentrated, aliquoted, and
stored at -80 ℃. (C). Flowchart of deamination assay on 3’ FAM labeled ssDNA substrate containing a
single target C. Briefly, various amount of purified proteins were incubated with 600nM FAM-labeled
ssDNA (5’-ATT TAT ATT ATT TAT TCA TAT TTA TAT TTA- FAM -3’) in deamination buffer
(HEPEs pH 6.5, 5 mM CaCl
2
, 1 mM DTT, 0.1 μg/ml RNase A) at 37 ℃ for 3 hrs. The reaction was
terminated by heating to 95 ℃ for 10 min. Then 2 units of UDG was added to the reaction mixture and
incubated for 1 hr at 37 ℃, followed by heating to 95℃ for 10 min to terminate the UDG reaction.
After incubating with 0.1M NaOH at 95 ℃ for 10 min, deamination products were separated on 20%
urea-PAGE, visualized with Molecular Imager FX (Bio-Rad), and quantified with Quantity One 1-D
analysis software (Bio-Rad). (D). A typical gel analysis of the deamination assay of the four purified
A3F proteins shown in panels-A, -B. Deaminase assay with purified A3A served as a positive control
for the 13 nt deamination product. (E, F). Dose-dependent deamination activity of wt A3F-fl and A3F-
fl E67A mutant (panel-E), and the bar representation of the deaminase activity (see Table 1) for the two

18
A3F constructs. Error bars represent standard deviation from the mean of three independent
experiments.

19
Table 2.1. Deaminase activity of purified proteins of various A3F constructs.

Constructs A3F-fl activity
(nM/µM/hr)
A3F-CD2 activity
(nM/µM/hr)
Wt 7.59 ± 0.981 0.501 ± 0.00434
E67A (in CD1) 5.89 ± 0.753 –––
E251A (in CD2) No –––
E67A/E251A (in CD1/CD2) No –––
W310A (in CD2) 3.23 ± 0.435 0.00478 ± 0.000427
D311A (in CD2) 5.89 ± 0.292 0.105 ± 0.0155
D311K (in CD2) 10.50 ± 0.663 0.609 ± 0.0439
T312P (in CD2) 7.32 ± 0.494 0.523 ± 0.0283
W126A (in CD1) 1.90 ± 0.344

Note: The deaminase activity data in this table were expressed as (Product (nM)/Enzyme
(µM)/Reaction time (hr)), which are obtained by linear regression fitting (using GraphPad Prism 6
software) based on the data at the beginning of linear range of the deaminase assay. Details for
deaminase reaction conditions and substrate used are described in Figure 3.1C legend.

2.1.2 The effect of CD2 loop7 mutations on A3F deaminase activity.
Since previous studies suggest a critical role of loop 7 for DNA binding and deaminase activity for
some APOBEC proteins, such as A3G-CD2, A3B-CD2, and A3A [29, 57-59], we want to assess the
functional role of specific residues on loop 7 of both the CD1 and CD2 of A3F. We made point
mutations around the tip of CD2 loop 7 on the constructs of both A3F-fl and A3F-CD2 alone, including
W310A, D311A, D311K, and T312P, which are located next to the Zn-coordinated active center
(Figure 2.2A, 2.2B). All constructs were expressed in E. coli and purified to near homogeneity (Figure
2.2C). For the four mutants in A3F-fl constructs, subsequent deamination assays revealed the W310A

20
mutant showed the most reduced activity over the tested protein concentration range (Figure 2.2D),
with about 57% reduction of deaminase activity when compared to wild type at the linear range (Fig.
2.2E, Table 2.1). The A3F-fl D311A and A3F-fl T312P mutants showed very little effect on deaminase
activity. Interestingly, the A3F-fl D311K mutant showed significantly increased activity, with
approximately 38% increased activity over wild type A3F-fl.

The purified wild type A3F-CD2 alone also showed obvious deamination activity (Figure 2.2F, 2.2G,
Table 2.1). However, deaminase activity of the A3F-CD2 construct by itself (~0.501 nM/µM/hr
(product/substrate/hr)) is more than 10-fold lower than that of A3F-fl and A3F-fl E67A (Table 2.1),
suggesting that the presence of CD1 on the N-terminus of CD2 greatly enhanced the activity of CD2.
For the four loop 7 mutations on this A3F-CD2 construct, we observed a similar trend of deaminase
activity changes as with the A3F-fl construct, with the exception of a dramatic activity drop on D311A
(Figure 2.2F, 2.2G, Table 2.1). Again, the mutant A3F-CD2 W310A showed the most significant
impact, reducing the deaminase activity to background level, followed by the A3F-CD2 D311A
mutant, which showed about an 80% reduction of deaminase activity compared to wild type A3F-CD2.
As with the full length construct, A3F-CD2 T312P had little effect on deamination activity, and quite
interestingly the D311K mutation once again showed an increased activity, with approximately 22%
higher activity than wild type A3F-CD2 (Figure 2.2G, Table 2.1). To summarize, the above result
showed that the four mutations of the three CD2 loop 7 residues showed similar effects on deaminase
activity in the context of either the A3F-fl or the A3F-CD2 construct. Generally speaking, the W310A
mutation had a significant negative impact on deaminase activity, whereas D311K had a positive
impact.

21

Figure 2.2. The effects of CD2 loop7 mutations on A3F activities. (A). One view of the A3F-CD2
crystal structures [40, 41, 60], showing the CD2 loop 7 of A3F-CD2 (PDB ID 3WUS) and the three
residues (in yellow sticks) mutated in this study. (B). Cartoon representation of the A3F-fl and A3F-
CD2 constructs containing mutations W310A, D311A, D311K, and T312P on CD2 (marked with a
color strip). (C) The SDS-PAGE analysis of the purified eight A3F proteins from E. coli expression.
The proteins are purified as described in Fu et al.[56] (see Figure 1B). (D, E). Dose-dependent
deamination activity of A3F-fl and the corresponding four CD2 loop 7 mutants (panel-D), and the bar
representation of the respective deaminase activity (Table 1) for the A3F-fl constructs (panel-E). The
deaminase assay conditions are described in Figure 1C legend. (F, G). Dose-dependent deamination
activity of A3F-CD2 and the corresponding four CD2 loop 7 mutants (panel-F), and the bar
representation of the respective deaminase activity (see Table 1) for the five A3F-CD2 constructs
(panel-G). Various amounts of proteins for each A3F-CD2 construct between 0-115 µM were
incubated with 600 nM ssDNA for 3 hrs at 37℃ at pH 6.5. (H, I). ssDNA binding curves of wt and
mutant proteins for A3F-fl (panel-H) and A3F-CD2 (panel-I) using native gel shift assay (see notes in
Table 3.2). The calculated Kd values for each protein construct were listed in Table 2. Error bars
represent standard deviation from the mean of three independent experiments.

22
2.1.3 The effect of CD2 loop7 mutations on substrate DNA binding:
In order to examine whether the CD2 loop 7 mutations on A3F-fl and A3F-CD2 have altered substrate
binding affinity, we performed the ssDNA binding assay for each protein construct using a gel shift
assay as described in previous studies [61, 62], and the result is shown in Figure 2H and 2I and Table 2.
For the A3F-fl constructs, the A3F-fl and A3F-fl D311A, D311K, and T312P constructs all had similar
binding affinity, with a Kd of around 1.0 µM (Table 2.2). However, the A3F-fl W310A construct
showed a greater than 3-fold reduction of the binding affinity compared to that of A3F-fl, with a Kd of
~3.5 µM (Table 2.2). A similar observation was observed among the A3F-CD2 constructs, where the
A3F-CD2 W310A mutant showed a much more dramatic reduction (over 10-fold) of DNA binding
affinity compared with that of wt A3F-CD2, with a Kd of ~152.5 µM (Table 2.2), whereas A3F-CD2
D311A showed about a 3-fold reduction. The other two A3F-CD2 mutants (D311K and T312P)
showed no obvious change in DNA binding affinity. Thus, the significant reduction of DNA binding
for W310A of A3F-CD2 and A3F-fl appears to correlate well with the reduced deaminase activity for
their respective mutant constructs. The fact that the W310A mutation in A3F-CD2 has a much more
dramatic effect in reducing DNA binding than the same mutation in A3F-fl suggests that presence of
CD1 in the A3F-fl construct can somehow compensate for the loss of DNA binding that is potentially
controlled by this tryptophan on CD2 loop 7.

However, when comparing the DNA binding between wild type A3F-fl and A3F-CD2 alone, the
binding affinity for A3F-CD2 (Kd of 12.8 µM) is about 13-fold lower than that for the A3F-fl (Kd of
1.1 µM) (Table 2), indicating that CD1 plays an important role in enhancing the DNA binding to the fl
A3F. Therefore, this reduced DNA binding affinity of the truncated construct A3F-CD2 again shows
correlation with the lowered deaminase activity compared with that of full length A3F, and similarly,
the lowered deaminase activity of the W310A mutants can be correlated to their decreased DNA
binding affinity compared with their wild type constructs.

23

Table 2.2. ssDNA (30 nt) binding affinity of purified proteins of various A3F constructs.

Constructs A3F-fl (µM) A3F-CD2 (µM) A3F-CD1 (µM)
wt 1.101±0.3354 12.80 ±2.028
0.146 ±0.0047
W310A 3.447 ±1.034 152.5 ±30.96
–––
D311A 1.796 ±0.4253 32.12 ±7.369
–––
D311K 1.099 ±0.3590 14.56 ±2.346
–––
T312P 0.917 ±0.2173 13.65 ±2.111
–––
W126A 3.931 ±0.4170 –––
NO

Note: DNA binding assay was performed using native gel shift. Increasing amount of A3F proteins
were incubated with 15 nM FAM-labeled ssDNA (5’-ATT TAT ATT ATT TAT TCA TAT TTA TAT
TTA- FAM -3’) at 25 ℃ for 5 min, followed by native gel electrophoresis for 100 min at 4 ℃. The
FAM-labeled DNA and its complex with the protein on the gel were then visualized with Molecular
Imager FX (Bio-Rad), and quantified with Quantity One 1-D analysis software (Bio-Rad). Based on the
ratios between free ssDNA and ssDNA-protein complex of each reaction, a protein concentration
dependent ssDNA binding curve was created by “one site specific binding” model from Prism 6.
Dissociation Constant (Kd) of each A3F construct was calculated using the same software.

2.1.4 Roles of CD1 loop 7 for A3F DNA binding and deaminase activities.
Since the CD2 loop-7 W310A mutation significantly impacted both DNA binding and deaminase
activity, we wondered whether the corresponding tryptophan on loop 7 of CD1, W126, has similar role
(Figure 2.3A). Interestingly, the loop 7 sequence of A3F-CD1 is very similar to that of A3F-CD2 as
well as A3G-CD1 (Figure 2.3B). We constructed a W126A mutant in A3F-fl (A3F-fl W126A) and in
A3F-CD1 alone (A3F-CD1 W126A) (Figure 2.3C), and performed an additional DNA gel shift assay.
The results demonstrated that, even though A3F-CD1 alone showed strong ssDNA binding, mutant
A3F-CD1 W126A had no detectable ssDNA binding under the same condition (Figure 2.3D, 2.3E,

24
Table 2.2), demonstrating that W126 on CD1 loop 7 plays an important role in binding DNA by the
A3F-CD1 construct.

Despite the complete loss of DNA binding by the A3F-CD1 W126A mutant, it is worth noting that the
wild type A3F-CD1 alone has a ssDNA binding affinity (Kd of 0.15 µM, Figure 3F, Table 2) that is
~85-fold higher than that of wild type A3F-CD2 alone (Kd of 12.8 µM), and ~7-fold higher than that of
wild type A3F-fl (1.1 µM) (Table 2). This result suggests that the three dimensional arrangement of
CD1-CD2 in the context of a full length A3F construct may result in loss of the DNA binding affinity
displayed by the CD1 domain alone, potentially through steric hindrances by the presence of CD2 that
effectively reduces DNA accessibility and binding to CD1.

Considering the undetectable DNA binding by A3F-CD1 W126A, we wondered whether full length
A3F with the CD1 W126A mutation (A3F-fl W126A) will have a DNA binding affinity comparable to
that of the A3F-CD2 construct alone. Interestingly, the result revealed that the A3F-fl W126A construct
displayed significantly stronger DNA binding than A3F-CD2 alone, with a Kd of ~3.9 µM (Figure
2.3G, Table 2). This is over 3-fold higher than that of A3F-CD2 alone, which had a Kd of 12.8 µM.
However, this DNA binding affinity of A3F FL W126A is still about 3-fold less than that of the wild
type A3F-fl (Kd of 1.1 µM) (Table 2.2). In other words, the A3F-fl W126A shows a DNA binding
affinity that is somewhere between the wild type A3F-fl and A3F-CD2 alone. Taken together, even
though the CD1 mutation W126A resulted in loss of DNA binding with the CD1 domain alone,
presence of the W126A mutant CD1 domain at the N-terminus of A3F-CD2 can still significantly
enhance the DNA binding compared to wild type CD2 alone. This suggests that the three dimensional
arrangement of CD1 and CD2 in the full length A3F construct may somehow increase the DNA

25
binding affinity of CD2; it is possible that positively charged surface areas formed between the two
domains may have additional impact beyond the observed contribution from loop 7 of the CD1 domain.

The deamination assay showed that the A3F-fl W126A construct had ~75% reduction of deaminase
activity compared with wild type A3F-fl (Fig. 2.3H, Table 2.1), indicating the important role of the
W126 residue on the loop 7 of CD1 for activity. On the other hand, the deamination activity of A3F-fl
W126A was about 4-fold higher than that of wild type A3F-CD2 alone (Figure 2.3H, Table 2.1), again
suggesting that the presence of the mutated CD1 on loop 7 W126A can still effectively enhance the
deaminase activity of CD2, likely benefitted from the enhanced DNA binding affinity facilitated by the
CD1 as a whole.

26

Figure 2.3. Roles of CD1 loop 7 W126 for A3F DNA binding and deaminase activities. (A). A view
of the A3F-CD1 structure that is modeled based on A3F-CD2 crystal structures [40, 41, 63], showing
CD1 loop 7 and W126 (in yellow sticks) mutated in this study. (B). Sequence alignment of loop 7
residues of both CD1 and CD2 from A3F and A3G, showing the close homology between A3G-CD1,
and both the CD1 and CD2 of A3F, with the conserved W among these three CD domains. (C).
Cartoon representation of the A3F-CD1 and A3F-CD1 W126A constructs, and the SDS-PAGE analysis
of the corresponding purified proteins. (D, E). Native gel shift assay of ssDNA binding of A3F-CD1
(panel-D) and A3F-CD1 W126A (panel-E) constructs (see notes in Table 2 for details). (F, G) The
ssDNA binding curves of wt A3F-CD1 (panel-F), and a comparison of binding curves for A3F-fl, A3F-
fl W126A, and A3F-CD2 (panel-G) (see notes in Table 2 for details). The calculated Kd values for
each construct were listed in Table 2. (H) The comparison of the deamination activity for A3F-fl, A3F-
fl W126A, and A3F-CD2 (also see Table 1). Error bars represent standard deviation from the mean of
three independent experiments.

27

2.2. Discussion
This is the first study of systematically mutating the predicted DNA binding surface residues on both
CD1 and CD2 domain of A3F and at the same time evaluating the importance of each domain and their
key surface residues to ssDNA binding and deamination activity.
The A3F protein expression and purification system used in this study offers a fast and easy way to
obtain recombinant proteins for in vitro biochemistry study. We concluded that only CD2 of A3F is
catalytically active. In a previous study from Yoshiyuki et al, the full-length A3F and its mutant
derivatives were fused with a C-terminal hemagglutinin tag and purified from 293T mammalian cells
[48]. To evaluate deaminase activity of each construct, the proteins were bound to protein-G-Sepharose
and incubated with the substrate ssDNA for 20 hours. However, only very weak deamination activities
(4 fold or less) were observed with the wild-type A3F and A3F E67A mutant (Hakata and Landau
2006) [48]. No activity was observed with A3F E251A or the double mutant (Hakata and Landau 2006)
[48]. Our study demonstrated similar but much convincing results that both A3F wt and A3F E67A had
strong initial deaminase activities, while the other two had no detectable deamination activity (Figure
2.1F).

We investigated four amino acid residues on loop 7 and demonstrated their critical roles in the
deamination activity of A3F. The amino acid residue W310 located in the CD2 domain loop 7 is
conserved in most APOBECs, which suggests the importance of this amino acid residue for APOBECs
activity. Here we show that both A3F FL W310A and A3F CD2 W310A had significantly reduced
initial deaminase activities, which were caused by a large decrease of ssDNA binding. The double ring
structure of Tryptophan is therefore likely to be critical for DNA substrate binding and stabilizing.

28
The amino acid residue D311 has been studied extensively. It was substituted to Y in 2005 [64], which
was found that the A3F D311Y mutant has an increased preference for G or C in the -1 position of the
tri-nucleotide motif (TGC/CCC) instead of the wt A3F preference TTC. In addition, the sequences of
the loop 7 of A3G CD1 and A3F CD2 are exactly the same except the residue right after this Aspartic
Acid, and previous research done by Hongzhan Xu et al. [65] in 2004 showed the substitution of D to K
can inhibit the interaction between A3G FL and Vif. Mariani et al. [66] have determined the properties
of A3G from African green monkey (AGM), rhesus macaque (MAC), and mouse. The results of these
studies indicate these A3G FL mutants that had Lysine instead of Aspartic Acid were able to inhibit
HIV-1 replication and thus exhibited HIV-1 Vif resistant. We constructed D311A and D311K mutants
to explore the importance of D311 residue and to use A3F CD2 as a model to understand the driving
force of K to D substitution during the evolution. Our results indicate that D311 is critical for
deamination activity for both A3F FL and CD2 alone (Figure 2.3E, 2.3G). However, D311K mutant
has a similar or even stronger initial deaminase activity and ssDNA binding ability, which implies there
may be other changes in RNA binding, other hotspot, or other regulation advantages for the single
amino acid substitution.

As indicated previously, there is only one amino acid difference on loop 7 between A3G CD1 and A3F
CD2, which is a Threonine in A3F CD2 and a Proline in A3G CD1. Since A3G CD1 has no
deamination activity while A3F CD2 does. we constructed a A3F T312P mutant aiming for studying
whether the T312P substitution caused loss of deamination ability of A3F. Our result indicated that the
T312P mutant had a similar deamination activity and ssDNA binding ability to those of wild-type
protein in both A3F FL and CD2 (Figure 2.3E, 2.3G). Therefore, the inactivity of A3G CD1 is not due
to this single amino acid difference in loop 7.

29
The four mutations had similar behaviors in both A3F CD2 and A3F FL, suggests A3F CD2 was still
the determinant of ssDNA target in the A3F FL. On the other hand, the deamination activity of A3F FL
wt was approximate 13 folds higher than that of A3F CD2 wt, and the ssDNA binding ability was
approximate 10 folds higher than that of A3F CD2 as well, which suggests the CD1 domain either
enhance the ssDNA binding ability or stability when A3F FL was bound to target ssDNA. The ssDNA
binding ability differences among wt and mutants in both A3F FL and A3F CD2 were consistent with
their initial deaminase activity results, which suggested these changes of initial deaminase activities
were according to the ssDNA binding abilities changing. However, such changes of ssDNA binding
had less effects in A3F FL with the help of CD1 domain.

The mutation W126A blocks the ssDNA binding ability of A3F CD1 dramatically, which made it a
suitable candidate to further explorer the function of CD1 domain (Figure 3.2D, E) in the context of the
full length protein. After we generated A3F FL with W126A, the initial deaminase activity of A3F FL
W126A was approximate 1/4 of A3F FL wt and 3 folds of A3F CD2 wt (Figure 3.2H). Similarly, the
ssDNA binding ability of A3F FL W126A was between that of A3F FL wt and A3F CD2 wt, which
was also approximate 1/4 ofA3F FL wt and 3 folds of A3F CD2 wt (Figure 3.2G). These results
suggest that both loop 7 of CD1 and CD2 are likely to be involved in stabilizeing ssDNA binding in
A3F FL; however, only CD2 loop 7 is fully functional to stabilize ssDNA in A3F FL W126A, which
leads to much weaker ssDNA binding and deamination ability. On the other hand, comparing with A3F
CD2 alone, in A3F FL W126A the mutated CD1 domain still contributed to ssDNA binding by
providing the positive charged surfaces, which may facilitate the recruitment of the negative charged
ssDNA substrate in the early stage of reaction. Interestingly, this W126A mutant was reported to result
in reduction in virion incorporation [67] and and loss of antiviral function of A3F FL [31]. Our results
of A3F FL W126A are consistant with both discoveries mentioned above. The reduction in virion

30
incorporation could be caused by dramatic decrease of interaction between virus ssRNA/ssDNA and
A3F FL W126A, and the loss of antiviral function of A3F FL W126A could be caused by significant
decrease of deamination activity and ssDNA binding ability of the mutant.

In summary, we report on the in vitro characterizations of DNA binding and deaminase activities using
purified wild type and various mutant A3F proteins from an E. coli expression system. We showed
that, despite the fact that CD1 is catalytically inactive and CD2 is the only active deaminase domain,
presence of both a wild type and mutant CD1 domain in the context of full length A3F can greatly
enhance the deaminase activity observed when compared to the second domain alone. This
enhancement of catalytic activity is likely through an increase of overall substrate DNA binding due to
the addition of the CD1 domain. We further showed that the loop 7 of both CD1 and CD2 of A3F are
important for mediating substrate ssDNA binding of the individual A3F domains and play an important
role in overall deamination activity of A3F.

2.3. Methods
2.3.1.Cloning, Expression and Protein construction
Wild type A3F FL, A3F CD1 and A3F CD2 and their corresponding mutants were constructed in the
pMal-c5X vector (New England Biolabs) to be expressed as N-terminal maltose binding protein (MBP)
fusion proteins. All clones were sequenced to confirm the correct sequences.
All the constructions were transformed into E. Coli strain C43 (GE health). cells and grew in 1L LB
media under 37 ℃. 0.3 mM IPTG were used to every 1L cells to induce the expression of target
proteins at 16 ℃ for 18h in the shaker incubator. Cell pellets were harvested and lysed by french press

31
with lysis buffer (50mM Tris-HCl pH 8.0, 150mM NaCl, 0.5mM TCEP), followed by centrifuged at
10000 rpm for 1h. The MBP fusion proteins were purified by passing through a column Amylose resin
(New England Biolabs), and washed by using 10 column volumes of wash buffer (50mM Tris-HCL pH
8.0, 1M NaCl, 0.5mM TCEP, 0.1ug/ml RNase A), then eluted with elution buffer (50mM Tris-HCl pH
8.0, 150mM NaCl, 0.5mM TCEP, 20mM Maltose). The elution was concentrated and aliquoted into
multiple small tubes, and stored at -80 ℃ for further experiments. Target proteins were confirmed by
SDS-PAGE gel, and their concentrations were determined by standard BSA protein on the same gel.

2.3.2.Deamination assay
Briefly, various amount of purified proteins were incubated with 5mM EDTA for 5min, and then
reacted with 600nM FAM-labeled ssDNA (5’-ATT TAT ATT ATT TAT TCA TAT TTA TAT TTA-
FAM -3’) in deamination buffer (HEPEs pH 6.5, 5 mM CaCl
2
, 1 mM DTT, 0.1 μg/ml RNase A) at 37
℃ for 3 hrs. The reaction was terminated by heating to 95 ℃ for 10 min. Then 2 units of UDG was
added to the reaction mixture and incubated for 1 hr at 37 ℃, followed by heating to 95℃ for 10 min
to terminate the UDG reaction. After incubating with 0.1M NaOH at 95 ℃ for 10 min, deamination
products were separated on 20% urea-PAGE, visualized with Molecular Imager FX (Bio-Rad), and
quantified with Quantity One 1-D analysis software (Bio-Rad). The deaminase activity data were
expressed as (Product (nM)/Enzyme (µM)/Reaction time (hr)), which are obtained by linear regression
fitting (using GraphPad Prism 6 software) based on the data at the beginning of linear range of the
deaminase assay. Details for deaminase reaction conditions and substrate used are described in Figure
2.1C legend.

32
2.3.3.EMSA assay
In the binding assay, titrated A3F proteins were reacted with 15nM FAM-DNA under room
temperature (25℃) for 5min. After mixed with 15% glycerol, each reaction was resolved on 1%
agarose gel for 100 min at 4℃, visualized with Molecular Imager FX (Bio-Rad), and quantified with
Quantity One 1-D analysis software (Bio-Rad). The free ssDNA will remain at the bottom of the gel,
while ssDNA-protein complex will shift to a higher position (Figure 2.4A, B). Based on the ratios
between free ssDNA and ssDNA-protein complex of each reaction, a protein concentration dependent
ssDNA binding curve was created by “one site specific binding” model from Prism 6. Dissociation
Constant (Kd) of each protein under the experimental condition was also calculated by the same
software.

2.3.4.DNA substrate
The DNA substrates we used for binding and deamination assay were designed on the basis of target
APOBEC hot spots. The sequence was “5’-ATT TAT ATT ATT TAT TCA TAT TTA TAT TTA-
FAM -3’ ”, labeled with FAM to 3’ end of the DNA substrate to detect specifically. The sequences
around the center are AT rich but unlikely to form secondary structure.

33

Figure 2.4 Gel images of ssDNA binding assay of A3F FL wt, CD2 wt and their loop 7 mutants.
Multiple protein-ssDNA complex bands could be seen on the gel.

34

Chapter 3: Divalent Ion Effects on Nine APOBEC Proteins.

Reproduced with permission by Qihan Chen, Yang Fu, Fumiaki Ito,
Jiang Gu and Xiaojiang Chen.
Submitted to Journal of Molecular Biology.

Contributions: Q.C. designed the project, analyzed biochemistry functions, and wrote the
manuscript, Y.F. offered A3B proteins, F.I. offered A3A proteins, J.G. offered A3H proteins and
X.S.C supervised the project.

35
3.1 Results
3.1.1. Effects of ions on deaminase activity of APOBEC proteins in the in vitro cytidine
deamination assay.
To determine the initial deaminase activity of these nine APOBEC proteins in the presence of metal
ions, various amount of both APOBEC wild types and their inactive mutant (see methods) were
incubated with target substrate ssDNA with 5 mM Ca
2+
, Mg
2+
, Cu
2+
and Zn
2+
.

The ion effects on the four APOBEC proteins, A3B, A3C, A3F and A3H, are shown in Figure 1, while
the results of the rest of the five APOBECs can be found in Supplementary Figure S3.1. To simplify
the comparison, the deamination activity of each APOBEC protein was summarized in Table 3.1 with
the activity of each protein in the absence of 5 mM divalent ions normalized to 100% . It was clear that
Ca
2+
and Mg
2+
enhanced the initial deaminase activities of most APOBEC proteins, while Zn
2+
and
Cu
2+
inhibited the initial deaminase activities of most APOBEC proteins. However, a unique response
of A3C with Zn
2+
was observed that showed about 5 folds higher deamination activity than that without
extra ions (Table 3.1).

36

Table 3.1. Comparison of deamination activities of nine APOBEC proteins with and without
addition of divalent ions.

no ion Ca
2+
Mg
2+
Zn
2+
Cu
2+
AID 100% ~100% ~100% ~60% ~50%
Apo1 100% ~100% ~130% ~50% ~30%
A3A 100% ~80% ~80% 0% 0%
A3B 100% ~100% ~100%
a
~90% ~80%
A3C 100% ~175% ~250% ~500% 0%
A3DE 100% ~100% ~250% ~70% 0%
A3F 100% ~200% ~300%
b
~30% ~100%
A3G 100% ~110% ~130% ~50% ~60%
A3H 100% ~200% ~200% 0% 0%

Note:
a
the background from the A3B FL inactive mutant with Mg
2+
was subtracted from the
deamination activity of A3B Fl wt.
b
ssDNA degradation was shown in A3F FL wt with Mg
2+
, which
implies the initial deaminase activity of A3F FL wt with Mg
2+
may not be accurate.

The results of A3B and its inactive mutant in the presence of different ions were shown on Figure 3.1A.
Compared with the positive control A3A, the 13 mer product was shown in all the reactions of A3B FL
wt with or without extra ions. However, the background 13 mer product was also detected in the
presence of Mg
2+
and Zn
2+
. Therefore, when we calculated the initial deaminase activity, the
background activity of inactive mutant was subtracted from the A3B FL wt activity to generate the
“sub” activity for each ion. There was no significant difference between initial deaminase activities
with or without Ca
2+
and Mg
2+
, while the activities with Cu
2+
and Zn
2+
were lower significantly than
that without ion (Figure 3.1B).

37
Since the optimal deamination pH of A3C was 5.5 in our preliminary tests, we measured the initial
deaminase activities of A3C with ions at pH 5.5 (Figure 3.1C). The initial deaminase activities of A3C
increased approximate 1.7 and 2.5 folds with Ca
2+
and Mg
2+
. To our surprise, Zn
2+
increased the A3C
initial deaminase activity ~5 folds as well, which indicated that the effect of Zn
2+
was quite different
with the other eight APOBEC proteins we tested. On the contrary, Cu
2+
inhibited the A3C initial
deaminase activity entirely (figure 3.1D).

Another double domains APOBEC protein, A3F, produced clear 13 mer products with and without
extra ions. The reactions of the inactive mutant with no ion, Ca
2+
and Cu
2+
were clean at 13 mer
position on the gel, while both 30 mer and 13 mer were degraded in the reactions of the inactive mutant
with Mg
2+
(Figure 3.1E). This is likely due to the presence of contaminated E. coli nucleases in the
protein samples. The reactions of A3F with Ca
2+
and Mg
2+
revealed ~2 folds and ~3 folds higher initial
deaminase activities than that with no extra ion, while the initial deaminase activity decreased ~3 folds
with Zn
2+
. However, there was no significant change when extra Cu
2+
was added (Figure 3.1F).

Based on our preliminary studies, pH 7.0 was the best working pH for A3H. Under the pH 7 condition,
compared with the positive control using A3A, the 13 mer products were shown in the reactions of
A3H with no ion, Ca
2+
and Mg
2+
(Figure 3.1G). The initial deaminase activities of A3H increased ~2
folds with both Ca
2+
and Mg
2+
very significantly. However, extra 5 mM Cu
2+
and Zn
2+
inhibited the
deamination activity of MBP-A3B FL wt completely within the tested protein concentration range
(Figure 3.1H).

38

Figure 3.1. Deamination activities of A3B, A3C, A3F and A3H with and without extra ions. (A).
Gel image showing the deamination activity of A3B FL wt and the inactive (E255A) mutant. 8.4uM
protein was incubated with 600nM ssDNA substrate with or without 5mM extra ion for 2 hrs at 37℃ at
pH 6.5. (B). The deamination activity of A3B FL wt with each ion was calculated based on dose
response experiments (Figure S3.2). Error bars represent the standard deviation from the mean of three
independent experiments. (C). Gel image showing the deamination activity of A3C wt and the inactive
(E68A) mutant. 0.22uM protein was incubated with 600nM ssDNA substrate with or without 5mM
extra ion for 2 hrs at 37℃ at pH 5.5. (D). The deamination activity of A3C wt with each ion was
calculated based on dose response experiments (Supplementary Figure S3.2). Error bars represent the
standard deviation from the mean of three independent experiments. (E). Gel image showing the
deamination activity of A3F FL wt and the inactive (E251A) mutant. 15uM protein was incubated with
600nM ssDNA substrate with or without 5mM extra ion for 3 hrs at 37℃ at pH 6.5. (F). The
deamination activity of A3F FL wt with each ion was calculated based on dose response experiments
(Supplementary Figure S3.2). Error bars represent standard deviation from the mean of three
independent experiments. (G). Gel image showing the deamination activity of A3H wt and the inactive
(E56A) mutant. 0.5uM protein was incubated with 600nM ssDNA substrate with or without 5mM extra
ion for 2 hrs at 37℃ at pH 7.0. (H). The deamination activity of A3H wt with each ion was calculated
based on dose response experiment (Supplementary Figure S3.2). Error bars represent standard
deviation from the mean of three independent experiments.

39

Figure S3.1. Deamination activities of AID, Apo1, A3A, A3D, and A3G with and without extra
ions. The deamination activities with each ion was calculated based on dose response experiment
(Figure S3.2). Error bars represent standard deviation from the mean of three independent experiments.

40

Figure S3.2. Dose response deamination activity curves of nine APOBEC proteins with and
without extra ions. The curves were generated based on the gel image of each APOBEC protein does
response experiment. Error bars represent standard deviation from the mean of three independent
experiments.

3.1.2 The amount of Ca
2+
affect A3F deamination activity and binding with ssDNA.
We investigated whether or not the concentration of ion is important for its effects on APOBEC
deamination activity and binding with ssDNA. Here we focused on A3F and Ca
2+
since Ca
2+
increased
the initial deaminase activity of A3F ~2 folds.

41
Various amounts of A3F proteins were incubated with the target ssDNA substrate with 1.25mM,
2.5mM, 5mM, 7.5mM and 10mM Ca
2+
at 37°C for 3 hours. A representative deamination gel was
displayed on Figure 3.2A. With the increasing amount of Ca
2+
in the reaction, there was no significant
difference in A3F initial deaminase activities from no ion to 2.5mM Ca
2+
. However, the initial
deaminase activities increased obviously when extra Ca
2+
concentration increased from 2.5 mM to 5
mM, and from 5 mM to 7.5 mM. Further increase of Ca
2+
from 7.5 mM to 10 mM did not show
obvious increase in the initial deaminase activity (Figure 3.2B).

Figure 3.2. Deamination activity and ssDNA binding of A3F FL wt in the presence of Ca
2+
. (A).
Gel image showing the deamination activity of A3F with various amounts of Ca
2+
. A3F FL wt at
concentration 15 µM, 7.5 µM, and 3.7 µM was incubated with 600 nM ssDNA substrate with 1.25
mM, 2.5 mM, 5 mM, 7.5 mM, and 10 mM Ca2+ for 3 hour at 37℃ at pH 6.5. (B). The initial
deaminase activity of A3F with different amounts of Ca
2+
was calculated based on dose response
experiment. Error bars represent standard deviation from the mean of three independent experiments. P
values were calculated based on unpaired double-tail t-test. (C). 0-15 µM A3F proteins were incubated
with 15 nM ssDNA substrate with 1.25 mM, 2.5 mM, 5 mM, 7.5 mM, and 10 mM Ca
2+
for 5 min at
25℃ and separated on 1% agarose gel. Error bars represent standard deviation from the mean of three
independent experiments. Kd of each condition was calculated based on the generated curve.

How did Ca
2+
enhance the deamination activity of A3F protein? To answer this question, we tested
ssDNA binding of A3F in the presence of different amount Ca
2+
(Figure 3.2C). The Kd of each group

42
was calculated based on the related curve. The results showed a clear trend that the more Ca
2+
was
added in the reaction, the lower Kd was observed. The binding between A3F and ssDNA substrate
increased from no ion to 1.25 mM Ca
2+
, 1.25 mM to 2.5 mM Ca
2+
and 2.5 mM to 5 mM Ca
2+
, and there
was no obvious increase from 5 mM to 7.5 mM Ca
2+
or 7.5 mM to 10 mM Ca
2+
. These results were
consistent with the observed initial deaminase activities of A3F wt with different amounts of Ca
2+
,
which suggested that Ca
2+
increased the initial deaminase activity of A3F by enhancing ssDNA binding
ability.

3.1.3 The unique effect of Zn
2+
on A3C
We wanted to test whether A3C had similar behavior as A3F in deamination activity and ssDNA
binding in the presence of Ca
2+
. By using the similar tests on A3F, the results of A3C demonstrated
similar trends in both initial deaminase activities and ssDNA binding abilities with Ca
2+
as those of
A3F (Figure 3.3A,3.3B,3.3C).

43

Figure 3.3. Deamination activity and ssDNA binding of A3C with Ca
2+
and Zn
2+
. (A). Gel image
showing the deamination activity of A3C wt with various amounts of Ca
2+
. A3C wt at concentration 56
nM, 28 nM, 14 nM, and 7 nM was incubated with 600 nM ssDNA substrate with 0 mM, 1.25 mM, 2.5
mM, 5 mM, 7.5 mM, and 10 mM Ca
2+
for 2 hour at 37 ℃ at pH 5.5. (B). The initial deaminase activity
of A3C with different amounts of Ca
2+
was calculated based on dose response experiments. Error bars
represent standard deviation from the mean of three independent experiments. P values were calculated
based on unpaired double-tail t-test. (C). 0-45 nM A3C proteins were incubated with 15 nM ssDNA
substrate with 1.25 mM, 2.5 mM, 5 mM, 7.5 mM, and 10 mM Ca
2+
for 5 min at 25 ℃ and separated on
1% agarose gel. Error bars represent standard deviation from the mean of three independent
experiments. Kd of each condition was calculated based on the generated curve. (D). Gel image
showing the deamination activity of A3C with various amounts of Zn
2+
. A3C at concentration 56 nM,
28 nM, 14 nM, and 7 nM was incubated with 600 nM ssDNA substrate with 0 mM, 1.25 mM, 2.5 mM,
5 mM, 7.5 mM, and 10 mM Ca
2+
for 2 hour at 37 ℃ at pH 5.5. (E). The initial deaminase activity of
A3C with different amounts of Zn
2+
was calculated based on dose response experiments. Error bars
represent standard deviation from the mean of three independent experiments. P values were calculated
based on unpaired double-tail t-test. (F). 0-220 nM A3C proteins were incubated with 15 nM ssDNA
substrate with 1.25 mM, 2.5 mM, 5 mM, 7.5 mM, and 10 mM Zn
2+
for 5 min at 25 ℃ and separated on
1% agarose gel. Error bars represent standard deviation from the mean of three independent
experiments. Kd of each condition was calculated based on the generated curve.

44

Interestingly, the change of initial deaminase activity with 5 mM Zn
2+
seemed different to the other
eight APOBEC proteins.

To investigate whether Zn
2+
affect the deamination activity of MBP-A3C wt in a different way, we first
tested the ssDNA binding ability of A3C with no extra ion, Ca
2+
, Mg
2+
and Zn
2+
by ssDNA gel shift
assay (Figure 3.4A). The Kd of A3C without any extra ion was around 6.4 nM, which was much lower
than that of A3F. The Kd of A3C with 5 mM Ca
2+
and Mg
2+
were both around 2 nM, which were only
1/3 of that without extra ion, indicating stronger ssDNA binding abilities. The ssDNA binding ability
of A3C wt with 5 mM Zn
2+
that demonstrated the strongest initial deaminase activity had a much
weaker ssDNA binding ability with the Kd around 65 nM, which was similar to that of A3F.

3.1.4 The amount of Zn
2+
affect A3C deamination activity and ssDNA binding in a different way.
It was interesting to explore why 5 mM Zn
2+
increased A3C initial deaminase activity with weaker
ssDNA binding ability. Similar to the test of A3C with Ca
2+
, various amount of A3C proteins were
incubated with target substrate ssDNA with no extra ion, 1.25 mM, 2.5 mM, 5 mM, 7.5 mM and 10
mM Zn
2+
at 37C for 2 hours. The gel was displayed as an example (figure 3.3D). Different to the
results of A3C with Ca
2+
, with the increasing amount of Zn
2+
in the reaction, the deamination activity
of A3C had an obvious peak when Zn
2+
concentration was 5 mM. The initial deaminase activities of
A3C increased from no extra ion to 1.25 mM and from 2.5 mM to 5 mM, and then decreased from 5
mM to 7.5 mM and 7.5 mM to 10 mM very significantly. There was no obvious change from extra 1.25
mM and 2.5 mM Zn
2+
(figure 3.3E).

45
To further confirm the ssDNA binding ability of A3C with Zn
2+
, we generated the binding curves
between A3C proteins and substrate ssDNA with various amounts of Zn
2+
from 0 mM, 1.25 mM, 2.5
mM, 5 mM, 7.5 mM to 10 mM (figure 3.3F). The result showed an exact opposite trend to that of A3C
initial deaminase activities with different amount of Zn
2+
. The highest Kd was around 65 nM , which
was shown when the extra Zn
2+
concentration was 5 mM. The Kd was decreased when the extra Zn
2+

concentration drop to 0 mM or raised as high as 10 mM. Those results clearly suggested that Zn
2+

enhanced A3C initial deaminase activity the most at a certain concentration around 5 mM. However,
such enhancement was not due to the stronger binding between A3C and ssDNA substrate, suggesting
that Zn
2+
may regulate A3C protein in other specific mechanism such as protein oligomerization or hot
spot recognition.

3.1.5 the extra ions did not change the binding between A3B and ssDNA obviously.
The change of A3B initial deaminase activities with extra ions were not obvious as the other two
domains APOBEC proteins A3F, which makes the reason interesting (figure 3.4B). Here we also tested
the binding abilities between A3B and ssDNA substrate with or without extra ions by EMSA (figure
3.4B). There was no obvious difference among the ssDNA binding abilities with extra 5 mM Ca
2+
, 5
mM Mg
2+
and without extra ion, which was consistent with the related initial deaminase activity data.
In addition, the Kd of binding with 5 mM Zn
2+
was slightly higher than that without extra ion, which
means the binding ability decreased a little bit. The data suggests that A3B was not as sensitive as other
APOBECs to extra ions in deamination activity and ssDNA binding.

46

Figure 3.4 ssDNA binding of A3C wt and A3B FL wt with and without extra ions. (A). 0-220 nM
A3C wt proteins were incubated with 15 nM ssDNA substrate without and with 5 mM extra Ca
2+
,
Mg
2+
, and Zn
2+
for 5 min at 25 ℃ and separated on 1% agarose gel. Error bars represent standard
deviation from the mean of three independent experiments. Kd of each condition was calculated based
on the generated curve. (B). 0-8.4 µM A3B FL wt proteins were incubated with 15 nM ssDNA
substrate without and with 5mM extra Ca
2+
, Mg
2+
, and Zn
2+
for 5 min at 25 ℃ and separated on 1%
agarose gel. Error bars represent standard deviation from the mean of three independent experiments.
Kd of each condition was calculated based on the generated curve.

3.2 Discussion
Previously, we tried to express and purify both A3F FL and the CD2 domain from E. coli. with a
Glutathione S-transferase (GST) tag or a 6His tag. However, only the CD2 domain itself was soluble
and active. These results indicated that the folding of the CD1 domain and the full length A3F are more
difficult than the CD2 domain alone in the E. coli expression system. We tried the refolding of the
insoluble A3F FL protein in vitro as well, which could make it close to the native structure that
confirmed by the circular dichroism spectroscopy. However, this refolded A3F protein had no
detectable deamination activity, which implies its lacking of certain cofactors such as ions or protein
modification. The same problem was also observed with most other APOBEC proteins. Later on,
several studies demonstrated a successful purification of soluble AID fused with Maltose Binding
Protein (MBP) [68, 69], offering a new strategy to solve the solubility problem. The MBP itself can

47
serve as a chaperone to help folding of the fused target protein [70], which may help APOBEC proteins
fold properly when they are expressed in E. coli. Our results demonstrated that it is feasible to obtain
the soluble APOBECs and their mutants fused with MBP at a large quantity and a high purity from E.
coli (Supplement Figure S3.3).

A3F is well known for its antiviral ability against HIV, and it is important to obtain a large amount of
pure protein with clean activity for structural and functional studies. We first tested the inactive
mutants to provide a negative control. However, degradation of the substrate was frequently observed,
likely due to contaminating nucleases, which was avoidable by addition of EDTA in the reaction. The
optimal concentration range of EDTA was determined to be between 5 to 10 mM, which caused no
detectable inhibition of deamination activity of APOBECs. Although we added up to 10mM extra ions
in the reaction, the real concentrations of free ions were much lower due to the existence of 5mM
EDTA. However, the trends of deaminaiton activities and ssDNA binding abilities with different
amounts of extra ions should be similar to real situation with much lower ion concentration at nM
range.

Since one of the A3G-CD2 crystal structure indicated an extra zinc ion and two extra magnesium ions
on the oligomerization interface [51], we were wondering whether those ions affect oligomer status of
APOBEC proteins or not. Based on the results from superdex 6 gel filtration column (GE Healthcare),
MBP fusion proteins purified from E. Coli. remained an oligomer around 1000 kDa though, and up to
10mM Ca
2+
or Mg
2+
did not dissociate it into small oligomer (data not shown). However, MALS of
APOBECs with different different amounts of divalent ions could be done to further confirm it in the
future.

48
The deamination activities of most APOBEC proteins demonstrated similar trends that 5 mM Ca
2+
and
Mg
2+
had positive effects on most of them, while 5 mM Zn
2+
or Cu
2+
had negative effects on most of
them (table 3.1). The ssDNA binding assay results of A3B, A3C and A3F suggested the mechanism of
Ca
2+
and Mg
2+
effects were due to the increase of ssDNA binding. On the contrary, The other two ions
seemed to inhibit the deamination activity of most APOBEC proteins due to the decrease of ssDNA
binding.

Further tests about A3F and A3C with different amounts of extra Ca
2+
indicated a clear correlation
between higher deamination activities and stronger ssDNA binding abilities (Figure 3.2, 3.3). The
divalent ion could work as a salt bridge to facilitate the wrapping of DNA on the surface of protein,
which was widely reported in the previous studies [71]. Based on the ssDNA shift gel, we did see more
than one shifted ssDNA band, which suggests there were more than one A3F protein binding on one
ssDNA substrate (figure S3.2). However, these patterns were similar with and without extra Ca
2+
,
implies that the oligomerization status did not change obviously with extra Ca
2+
.

On the other hand, A3C with Zn
2+
seemed to be an obvious exception comparing with other single
domain or double domains APOBEC proteins. Although Ca
2+
and Mg
2+
still enhanced the initial
deaminase activities of A3C around 2 folds like most other APOBEC proteins, the initial deaminase
activity with Zn
2+
was higher than that with no extra ion approximate 5 folds (figure 3.3). To our
surprise, the curve of initial deaminase activities was a peak instead of an “S” shape curve with the
increase of Zn
2+
, whose highest point was shown when the concentration of Zn
2+
was 5 mM. Opposite
to the change of A3C initial deaminase activity, the ssDNA binding ability was actually displayed as a
valley shape with the lowest point as 5 mM Zn
2+
as well (figure 3.3E). This data suggested that Zn
2+

regulate A3C in a different way to those with other ions. It may loose the strong binding between
APOBEC protein and ssDNA to facilitate the protein sliding-targeting process since the Kd of A3C

49
was extremely low. However, if the concentration of Zn
2+
is too high, the inhibition effect on A3C as
well as other APOBEC proteins is over the facilitation effect. In the previous studies, Kd of A3G was
around 76 nM measured by gel shift assay [72], and Kd of A3A was approximate 330 nM measured by
anisotropy [73]. Although the conditions were different, A3C may have an even stronger ssDNA
binding ability whose Kd was around 6 nM, which could be an very interesting direction to explore the
related biological function in the future.

In summary, we report on the in vitro characterizations of ssDNA binding and deaminase activities
using purified nine wild type and inactive mutant APOBEC proteins from an E. coli expression system.
We showed that, four divalent ions, Ca, Mg, Zinc, and Cu can affect ssDNA binding ability and
deaminase activities of APOBECs. Most of those enhancement and inhibition effects of catalytic
activity by extra divalent ions are likely through an increase or decrease of overall substrate ssDNA
binding. We further showed that Zn could increase A3C deaminase activity up to 5 folds at certain
concentration points, but not due to the stronger binding with substrate ssDNA like others.

3.3 Methods
3.3.1. Cloning, expression and Protein construction
AID wt and its inactive mutant (E58A), Apo1 wt and its inactive mutant (E63A), A3B FL wt and its
inactive mutant (E255A), A3C and its inactive mutant (E68A), A3DE and its inactive mutant (E264A),
A3F FL wt and its inactive mutant (E251A), A3G and its inactive mutant (E259A), A3H wt and its
inactive mutant (E56A) DNA coding sequences were synthesized and codon optimized for E. Coli
expression. All three were constructed in the pMal-c5X vector (New England Biolabs) to be expressed
as N-terminal maltose binding protein (MBP) fusion proteins. A3A wt and its inactive mutant (EA)
DNA coding sequences were synthesized and codon optimized for E. Coli expression with a C-terminal

50
His tag, then constructed in the pET-28a(+) vector (Novagen). All clones were sequenced to confirm
the correct sequences.
All the constructions were transferred into E. Coli strain C43 (GE health). cells and grew in 1L LB
matrix under 37 ℃. 0.3mM IPTG were used to every 1L cells to induce the expression of target
proteins at 16 ℃ for 18h in the shaker incubator. Cell pellets were harvested and lysed by french press
with lysis buffer (50 mM Tris-HCl pH 8.0, 150 mM NaCl, 0.5 mM TCEP), followed by centrifuged at
10000 rpm for 1h. The MBP fusion proteins were purified by passing through a column Amylose resin
(New England Biolabs), and washed by using 10 column volumes of wash buffer (50 mM Tris-HCL
pH 8.0, 1 M NaCl, 0.5 mM TCEP, 0.1ug/ml RNase A), then eluted with elution buffer (50 mM Tris-
HCl pH 8.0, 150 mM NaCl, 0.5 mM TCEP, 20 mM Maltose). The His tag fusion protein was purified
by Nickel resin column (Qiagen), and washed by using 10 column volumes of wash buffer (50 mM
Tris-HCl, 150 mM NaCl, 0.5 mM TCEP, 50 mM imidazole), then eluted with elution buffer (50 mM
Tris-HCl pH 8.0, 150 mM NaCl, 0.5 mM TCEP, 500 mM Imidazole) . The elution was concentrated
and aliquoted into multiple small tubes, and stored at -80 ℃ for further experiments. Target proteins
were proved by SDS-PAGE gel, and their concentrations were determined by standard BSA protein on
the same gel (Figure S3.3).

3.3.2. Deamination assay
APOBEC proteins were incubated with 5 mM EDTA for 5 min, and then reacted with 600 nM FAM-
labeled ssDNA substrate (synthesized by Integrated DNA Technologies) in deamination buffer (50 mM
HEPEs pH 7.0/HEPEs pH 6.5/Mes pH 5.5, 1mM DTT, 0.1ug/ml RNase A). Reactions were incubated
at 37 ℃ for designed time lengths and terminated by heating to 95℃ for 10 min, followed by 1h
incubation with 2 units of UDG at 37 ℃ and terminated by heating to 95℃ for 10 min again. After

51
incubating with 0.1 M NaOH at 95 ℃ for 10 min, deamination products were separated on 20% urea-
PAGE, visualized with Molecular Imager FX (Bio-Rad), and quantified with Quantity One 1-D
analysis software (Bio-Rad). Based on the ratio between the amounts of ssDNA substrate and
deamination product, protein concentration dependent ssDNA deamination curve was created by
“Mechaelis-Menten kinetic” model from Prism 6. Deamination activity, measured as nM of substrate
deaminated per µM of enzyme per hour, was calculated from the linear part at the beginning of
deamination curve.

3.3.3. EMSA assay
In the binding assay, titrated A3F proteins were reacted with 15 nM FAM-DNA under room
temperature (25℃) for 5 min. After mixed with 15% glycerol, each reaction was resolved on 1%
agarose gel for 100 min at 4℃, visualized with Molecular Imager FX (Bio-Rad), and quantified with
Quantity One 1-D analysis software (Bio-Rad). The free ssDNA will remain at the bottom of the gel,
while ssDNA-protein complex will shift to a higher position (Figure S3.4). Based on the ratios between
free ssDNA and ssDNA-protein complex of each reaction, a protein concentration dependent ssDNA
binding curve was created by “one site specific binding” model from Prism 6. Dissociation Constant
(Kd) of each protein under the experimental condition was also calculated by the same software.

3.3.4. DNA substrate
The DNA substrates we used for binding and deamination assay were designed on the basis of target
APOBEC hot spots from our preliminary tests. The sequence for A3DE was “5’-fam-ATT TAT ATT
ATT TAT ACA TAT TTA TAT TTA-3’ ”, and the sequence for A3G was “5’-fam-ATT TAT ATT

52
ATT TAT CCC TAT TTA TAT TTA-3’ ”. The sequences for rest of APOBECs were “5’-ATT TAT
ATT ATT TAT TCA TAT TTA TAT TTA- FAM -3’ ”The florescence label FAM was linked to either
5’ end or 3’ end of the DNA substrate to detect specifically. The sequences around the center are AT
rich but unlikely to form secondary structure.

Figure S3.3 Coomassie gel images of nine APOBEC proteins and their inactive
mutants.

53

Figure S3.4 Gel images of EMSA assay of A3F FL wt with various amounts of Ca
2+
.
Multiple protein-ssDNA complex bands could be seen on the gel.

54
Bibliography
1. Jarmuz, A., et al., An anthropoid-specific locus of orphan C to U RNA-editing enzymes
on chromosome 22. Genomics, 2002. 79(3): p. 285-96.
2. Bransteitter, R., C. Prochnow, and X.S. Chen, The current structural and functional
understanding of APOBEC deaminases. Cell Mol Life Sci, 2009. 66(19): p. 3137-47.
3. Conticello, S., The AID/APOBEC family of nucleic acid mutators. Genome Biology,
2008. 9(6): p. 229.
4. Harris, R., et al., DNA deamination mediates innate immunity to retroviral infection.
Cell, 2003. 113: p. 803 - 809.
5. Yu, X.F., Innate cellular defenses of APOBEC3 cytidine deaminases and viral counter-
defenses. Curr Opin HIV AIDS, 2006. 1(3): p. 187-93.
6. Prochnow, C., R. Bransteitter, and X.S. Chen, APOBEC deaminases-mutases with
defensive roles for immunity. Sci China C Life Sci, 2009. 52(10): p. 893-902.
7. Harris, R.S. and M.T. Liddament, Retroviral restriction by APOBEC proteins. Nat Rev
Immunol, 2004. 4(11): p. 868-877.
8. Duggal, N.K. and M. Emerman, Evolutionary conflicts between viruses and restriction
factors shape immunity. Nat Rev Immunol, 2012. 12(10): p. 687-695.
9. Moris, A., S. Murray, and S. Cardinaud, AID and APOBECs span the gap between
innate and adaptive immunity. Frontiers in Microbiology, 2014. 5: p. 534.
10. Koito, A. and T. Ikeda, Intrinsic immunity against retrotransposons by APOBEC
cytidine deaminases. Front Microbiol, 2013. 4.
11. Thielen, B.K., et al., Innate Immune Signaling Induces High Levels of TC-specific
Deaminase Activity in Primary Monocyte-derived Cells through Expression of
APOBEC3A Isoforms. The Journal of Biological Chemistry, 2010. 285(36): p. 27753-
27766.
12. Russell, R.A., et al., Distinct Domains within APOBEC3G and APOBEC3F Interact with
Separate Regions of Human Immunodeficiency Virus Type 1 Vif. Journal of Virology,
2009. 83(4): p. 1992-2003.
13. Farrow, M.A. and A.M. Sheehy, Vif and Apobec3G in the innate immune response to
HIV: a tale of two proteins. Future Microbiol, 2008. 3(2): p. 145-54.
14. Russell, R.A. and V.K. Pathak, Identification of Two Distinct Human Immunodeficiency
Virus Type 1 Vif Determinants Critical for Interactions with Human APOBEC3G and
APOBEC3F. Journal of Virology, 2007. 81(15): p. 8201-8210.
15. Takaori-Kondo, A., APOBEC Family Proteins: Novel Antiviral Innate Immunity.
International Journal of Hematology. 83(3): p. 213-216.
16. Cho, S.-J., et al., APOBEC3F and APOBEC3G mRNA Levels Do Not Correlate with
Human Immunodeficiency Virus Type 1 Plasma Viremia or CD4(+) T-Cell Count.
Journal of Virology, 2006. 80(4): p. 2069-2072.
17. Zheng, Y.H., et al., Human APOBEC3F is another host factor that blocks human
immunodeficiency virus type 1 replication. J Virol, 2004. 78(11): p. 6073-6.
18. Burns, M.B., et al., APOBEC3B is an enzymatic source of mutation in breast cancer.
Nature, 2013. 494(7437): p. 366-70.
19. Burns, M.B., N.A. Temiz, and R.S. Harris, Evidence for APOBEC3B mutagenesis in
multiple human cancers. Nat Genet, 2013. 45(9): p. 977-983.

55
20. Nik-Zainal, S., et al., Association of a germline copy number polymorphism of
APOBEC3A and APOBEC3B with burden of putative APOBEC-dependent mutations in
breast cancer. Nat Genet, 2014. 46(5): p. 487-491.
21. Roberts, S.A., et al., An APOBEC cytidine deaminase mutagenesis pattern is
widespread in human cancers. Nat Genet, 2013. 45(9): p. 970-976.
22. Conticello, S.G., Creative deaminases, self-inflicted damage, and genome evolution.
Annals of the New York Academy of Sciences, 2012. 1267(1): p. 79-85.
23. Guo, Junjie U., et al., Hydroxylation of 5-Methylcytosine by TET1 Promotes Active DNA
Demethylation in the Adult Brain. Cell. 145(3): p. 423-434.
24. Nik-Zainal, S., et al., Mutational Processes Molding the Genomes of 21 Breast
Cancers. Cell. 149(5): p. 979-993.
25. Conticello, S., R. Harris, and M. Neuberger, The Vif protein of HIV triggers degradation
of the human antiretroviral DNA deaminase APOBEC3G. Curr Biol, 2003. 13: p. 2009 -
2013.
26. Mariani, R., et al., Species-specific exclusion of APOBEC3G from HIV-1 virions by Vif.
Cell, 2003. 114(1): p. 21-31.
27. Sheehy, A.M., N.C. Gaddis, and M.H. Malim, The antiretroviral enzyme APOBEC3G is
degraded by the proteasome in response to HIV-1 Vif. Nat Med, 2003. 9(11): p. 1404-
1407.
28. Wang, X., et al., Biochemical Differentiation of APOBEC3F and APOBEC3G Proteins
Associated with HIV-1 Life Cycle. The Journal of biological chemistry, 2007. 282(3): p.
1585-1594.
29. Liddament, M.T., et al., APOBEC3F properties and hypermutation preferences indicate
activity against HIV-1 in vivo. Curr Biol, 2004. 14(15): p. 1385-91.
30. Wang, X., et al., Human APOBEC3F incorporation into human immunodeficiency virus
type 1 particles. Virus Research, 2014. 191: p. 30-38.
31. Song, C., et al., Signals in APOBEC3F N-terminal and C-terminal deaminase domains
each contribute to encapsidation in HIV-1 virions and are both required for HIV-1
restriction. J Biol Chem, 2012. 287(20): p. 16965-74.
32. Zennou, V., et al., APOBEC3G Incorporation into Human Immunodeficiency Virus
Type 1 Particles. Journal of Virology, 2004. 78(21): p. 12058-12061.
33. Wang, T., et al., Interaction with 7SL RNA but Not with HIV-1 Genomic RNA or P
Bodies Is Required for APOBEC3F Virion Packaging. Journal of Molecular Biology,
2008. 375(4): p. 1098-1112.
34. Navarro, F., et al., Complementary function of the two catalytic domains of
APOBEC3G. Virology, 2005. 333(2): p. 374-86.
35. Li, X., et al., Functional analysis of the two cytidine deaminase domains in
APOBEC3G. Virology, 2011. 414(2): p. 130-136.
36. Wang, T., et al., 7SL RNA Mediates Virion Packaging of the Antiviral Cytidine
Deaminase APOBEC3G. Journal of Virology, 2007. 81(23): p. 13112-13124.
37. Apolonia, L., et al., Promiscuous RNA Binding Ensures Effective Encapsidation of
APOBEC3 Proteins by HIV-1. PLoS Pathogens, 2015. 11(1): p. e1004609.
38. Liu, B., et al., Regulation of Apobec3F and Human Immunodeficiency Virus Type 1 Vif
by Vif-Cul5-ElonB/C E3 Ubiquitin Ligase. Journal of Virology, 2005. 79(15): p. 9579-
9587.

56
39. Smith, J.L. and V.K. Pathak, Identification of Specific Determinants of Human
APOBEC3F, APOBEC3C, and APOBEC3DE and African Green Monkey APOBEC3F
That Interact with HIV-1 Vif. Journal of Virology, 2010. 84(24): p. 12599-12608.
40. Nakashima, M., et al., Structural Insights into HIV-1 Vif-APOBEC3F Interaction. J Virol,
2015. 90(2): p. 1034-47.
41. Siu, K.K., et al., Structural determinants of HIV-1 Vif susceptibility and DNA binding in
APOBEC3F. Nat Commun, 2013. 4: p. 2593.
42. Holmes, R.K., et al., APOBEC3F can inhibit the accumulation of HIV-1 reverse
transcription products in the absence of hypermutation. Comparisons with
APOBEC3G. J Biol Chem, 2007. 282(4): p. 2587-95.
43. Aydin, H., Matthew W. Taylor, and Jeffrey E. Lee, Structure-Guided Analysis of the
Human APOBEC3-HIV Restrictome. Structure. 22(5): p. 668-684.
44. Stenglein, M.D. and R.S. Harris, APOBEC3B and APOBEC3F inhibit L1
retrotransposition by a DNA deamination-independent mechanism. J Biol Chem, 2006.
281(25): p. 16837-41.
45. Bélanger, K., et al., Binding of RNA by APOBEC3G controls deamination-independent
restriction of retroviruses. Nucleic Acids Research, 2013. 41(15): p. 7438-7452.
46. Albin, J.S., W.L. Brown, and R.S. Harris, Catalytic Activity of APOBEC3F Is Required
for Efficient Restriction of Vif-deficient Human Immunodeficiency Virus. Virology, 2014.
0: p. 49-54.
47. Mbisa, J.L., W. Bu, and V.K. Pathak, APOBEC3F and APOBEC3G inhibit HIV-1 DNA
integration by different mechanisms. J Virol, 2010. 84(10): p. 5250-9.
48. Hakata, Y. and N.R. Landau, Reversed functional organization of mouse and human
APOBEC3 cytidine deaminase domains. J Biol Chem, 2006. 281(48): p. 36624-31.
49. Holden, L.G., et al., Crystal structure of the anti-viral APOBEC3G catalytic domain and
functional implications. Nature, 2008. 456(7218): p. 121-4.
50. Newman, E.N., et al., Antiviral function of APOBEC3G can be dissociated from cytidine
deaminase activity. Curr Biol, 2005. 15(2): p. 166-70.
51. Shandilya, S.M., et al., Crystal structure of the APOBEC3G catalytic domain reveals
potential oligomerization interfaces. Structure, 2010. 18(1): p. 28-38.
52. Li, M., et al., First-in-class small molecule inhibitors of the single-strand DNA cytosine
deaminase APOBEC3G. ACS Chem Biol, 2012. 7(3): p. 506-17.
53. Bohn, M.-F., et al., The ssDNA Mutator APOBEC3A Is Regulated by Cooperative
Dimerization. Structure. 23(5): p. 903-911.
54. Chelico, L., et al., A model for oligomeric regulation of APOBEC3G cytosine
deaminase-dependent restriction of HIV. J Biol Chem, 2008. 283(20): p. 13780-91.
55. Domanico, P.L. and Y.C. Tse-Dinh, Mechanistic studies on E. coli DNA topoisomerase
I: Divalent ion effects. Journal of Inorganic Biochemistry, 1991. 42(2): p. 87-96.
56. Fu, Y., et al., DNA cytosine and methylcytosine deamination by APOBEC3B:
enhancing methylcytosine deamination by engineering APOBEC3B. Biochemical
Journal, 2015. 471(Pt 1): p. 25-35.
57. Logue, E.C., et al., A DNA Sequence Recognition Loop on APOBEC3A Controls
Substrate Specificity. PLoS ONE, 2014. 9(5): p. e97062.
58. Rathore, A., et al., The Local Dinucleotide Preference of APOBEC3G Can Be Altered
from 5′-CC to 5′-TC by a Single Amino Acid Substitution. Journal of Molecular Biology,
2013. 425(22): p. 4442-4454.

57
59. Wang, M., C. Rada, and M.S. Neuberger, Altering the spectrum of immunoglobulin V
gene somatic hypermutation by modifying the active site of AID. The Journal of
Experimental Medicine, 2010. 207(1): p. 141-153.
60. Bohn, M.F., et al., Crystal structure of the DNA cytosine deaminase APOBEC3F: the
catalytically active and HIV-1 Vif-binding domain. Structure, 2013. 21(6): p. 1042-50.
61. Mitra, M., et al., Structural determinants of human APOBEC3A enzymatic and nucleic
acid binding properties. Nucleic Acids Research, 2014. 42(2): p. 1095-1110.
62. Byeon, I.-J.L., et al., NMR structure of human restriction factor APOBEC3A reveals
substrate binding and enzyme specificity. Nature communications, 2013. 4: p. 1890-
1890.
63. Bohn, M.-F., et al., Crystal structure of the DNA cytosine deaminase APOBEC3F: the
catalytically active and HIV-1 Vif-binding domain. Structure, 2013. 21(6): p. 1042-1050.
64. Teng, B., C. Burant, and N. Davidson, Molecular cloning of an apolipoprotein B
messenger RNA editing protein. Science, 1993. 260: p. 1816 - 1819.
65. Xu, H., et al., A single amino acid substitution in human APOBEC3G antiretroviral
enzyme confers resistance to HIV-1 virion infectivity factor-induced depletion. Proc Natl
Acad Sci U S A, 2004. 101(15): p. 5652-7.
66. Mariani, R., et al., Species-specific exclusion of APOBEC3G from HIV-1 virions by Vif.
Cell, 2003. 114: p. 21 - 31.
67. Wang, T., et al., Interaction with 7SL RNA but not with HIV-1 genomic RNA or P bodies
is required for APOBEC3F virion packaging. J Mol Biol, 2008. 375: p. 1098 - 1112.
68. Coker, H.A. and S.K. Petersen-Mahrt, The nuclear DNA deaminase AID functions
distributively whereas cytoplasmic APOBEC3G has a processive mode of action. DNA
Repair, 2007. 6(2): p. 235-243.
69. Kohli, R.M., et al., A portable hot spot recognition loop transfers sequence preferences
from APOBEC family members to activation-induced cytidine deaminase. J Biol Chem,
2009. 284(34): p. 22898-904.
70. Bach, H., et al., Escherichia coli maltose-binding protein as a molecular chaperone for
recombinant intracellular cytoplasmic single-chain antibodies1. Journal of Molecular
Biology, 2001. 312(1): p. 79-93.
71. Saecker, R.M. and M.T. Record Jr, Protein surface salt bridges and paths for DNA
wrapping. Current Opinion in Structural Biology, 2002. 12(3): p. 311-319.
72. Iwatani, Y., et al., Biochemical activities of highly purified, catalytically active human
APOBEC3G: correlation with antiviral effect. J Virol, 2006. 80(12): p. 5992-6002.
73. Pham, P., et al., A biochemical analysis linking APOBEC3A to disparate HIV-1
restriction and skin cancer. J Biol Chem, 2013. 288(41): p. 29294-304.

Asset Metadata

Creator Chen, Qihan (author)

Core Title Biochemical studies of APOBEC protein family

Contributor Electronically uploaded by the author (provenance)

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Molecular Biology

Publication Date 04/27/2018

Defense Date 02/02/2016

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag deamination activity,ion effect,loop 7,OAI-PMH Harvest,ssDNA binding

Format application/pdf (imt)

Language English

Advisor Chen, Xiaojiang (committee chair), Chen, Lin (committee member), Qin, Peter (committee member)

Creator Email lyonchenqihan@hotmail.com,qihanche@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-244862

Unique identifier UC11276348

Identifier etd-ChenQihan-4383.pdf (filename),usctheses-c40-244862 (legacy record id)

Legacy Identifier etd-ChenQihan-4383.pdf

Dmrecord 244862

Document Type Dissertation

Format application/pdf (imt)

Rights Chen, Qihan

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Abstract (if available)

Abstract APOBEC3F (A3F) is a member of the apolipoprotein B mRNA-editing enzyme catalytic polypeptide-like (APOBEC) family of proteins that can deaminate cytosine (C) to uracil (U) on nucleic acids. A3F is one of the four APOBEC members with two Zn-coordinated homologous cytosine deaminase domains (CDs), with the others being A3B, A3DE and A3G. ❧ Here we report the first comprehensive in vitro characterization of DNA binding and deaminase activities using purified wild type (wt) and various mutant proteins of A3F from an Escherichia coli expression system. We show that, even though the N-terminal CD1 domain is catalytically inactive and the C-terminal CD2 domain is the active deaminase domain, presence of CD1 on the N-terminus of CD2 enhances the deaminase activity by over an order of magnitude. This enhancement of CD2 catalytic activity is mainly through the increase of substrate ssDNA binding by the N-terminal CD1 domain. We further show that the loop 7 of both CD1 and CD2 of A3F plays an important role for ssDNA binding for each individual domain as well as for the deaminase activity of CD2 domain in the full length A3F. ❧ The APOBEC family consists of eleven known members. Nine of them have been shown to deaminate cytosine (C) to uracil (U) on nucleic acid. The regulation mechanism of APOBEC proteins is still unclear. Here we demonstrated four divalent ion effects (Ca²⁺, Mg²⁺, Zn²⁺, and Cu²⁺) on nine APOBEC proteins. With the data of ssDNA binding ability, we revealed that Ca²⁺ and Mg²⁺ have positive effects on deamination activity due to the increase of ssDNA binding, while Zn²⁺ and Cu²⁺ have negative effects on deamination activity due to the decrease of ssDNA binding. In addition, we also showed Zn²⁺ had a special effect on A3C comparing with other APOBEC proteins and ions.