Article

ISSN 0006-2979, Biochemistry (Moscow), 2024, Vol. 89, No. 4, pp. 653-662 © Pleiades Publishing, Ltd., 2024.

653

Towards Development of the 4C-Based Method Detecting

Interactions of Plasmid DNA with Host Genome

Alexandra P. Yan

1,2,a

*, Paul A. Salnikov

1,2

, Maria M. Gridina

1,2

Polina S. Belokopytova

1,2

, and Veniamin S. Fishman

1,2,b

Institute of Cytology and Genetics, Siberian Branch of the Russian Academy of Sciences,

630090 Novosibirsk, Russia

Novosibirsk State University, 630090 Novosibirsk, Russia

e-mail: a.yan@g.nsu.ru

e-mail: minja-f@yandex.ru

Alexandra Yan https://orcid.org/0000-0003-0305-0612

Paul Salnikov https://orcid.org/0000-0001-9470-7178

Maria Gridina https://orcid.org/0000-0002-7972-5949

Polina Belokopytova https://orcid.org/0000-0003-1390-7341

Veniamin Fishman https://orcid.org/0000-0002-5573-3100

Received October 11, 2023

Revised February 1, 2024

Accepted March 2, 2024

Abstract— Chromosome conformation capture techniques have revolutionized our understanding of chromatin

architecture and dynamics at the genome-wide scale. In recent years, these methods have been applied to a diverse

array of species, revealing fundamental principles of chromosomal organization. However, structural organiza-

tion of the extrachromosomal entities, like viral genomes or plasmids, and their interactions with the host genome,

remain relatively underexplored. In this work, we introduce an enhanced 4C-protocol tailored for probing plasmid

DNA interactions. We design specific plasmid vector and optimize protocol to allow high detection rate ofcontacts

between the plasmid and host DNA.

DOI: 10.1134/S0006297924040059

Keywords: chromatin interactions, 4C, plasmid, extrachromosomal sequences, plasmid-genome interactions, chro-

mosome conformation capture, transfection

* To whom correspondence should be addressed.

INTRODUCTION

Recent advances in biochemistry, molecular biol-

ogy, and next-generation sequencing facilitated devel-

opment of highly efficient methods for probing inter-

phase chromatin architecture. It has become evident

that organization of the vertebrate genomes is influ-

enced by at least two distinct mechanisms. The first in-

volves interplay between the CTCF and cohesin factors

in cis [1], while the second pertains to organization of

the genomic compartments through phase separation,

mediated by interactions of chromatin proteins in cis

and trans [2]. Despite a few exceptions [3, 4], these

mechanisms typically operate in tandem within most

vertebrate cells. This combined function makes it dif-

ficult to distinguish between the independent factors

associated with each mechanism. This poses a partic-

ular challenge for studying compartment formation

through phase separation, because the generic term

“compartment” can be attributed to various nuclear

partitions, each characterized by unique chromatin

content and formed by a specific biochemical mecha-

nism. Thus, sequence determinants, proteins involved,

and additional aspects of chromatin compartmentaliza-

tion phenomenon are yet to be fully elucidated [5, 6].

Another significant challenge within the field of

3D-genomics is exploration of variability in the chro-

matin architectures across the tree of life. While evo-

lution of the genome architecture in vertebrates [7]

and some invertebrate groups [8] has been relatively

YAN et al.654

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

well-studied, data for other species remain limited.

Furthermore, there is a lack of information regard-

ing nuclear organization of non-genomic and exoge-

nous sequences, including mitochondrial and plastid

DNA, viral genomes, plasmids, mobile elements, and

so forth. Nevertheless, existing evidence indicates that

spatial packaging of these DNA molecules plays a cru-

cial role in their function, and studying it can uncover

important knowledge about the general principles of

chromatin biology.

Viral DNA within the nuclei serves as a prominent

example of exogenous DNA sequence. For example, it

has been established that the viral DNA of herpes B vi-

rus (HBV), characterized as episomal covalently closed

circular DNA, primarily contacts the A-compartment

of the genome. These contacts include promoters, en-

hancers, and CpG islands of the promoters of highly

expressed genes, including those engaged in the cellu-

lar response to infection [9]. Conversely, the transcrip-

tionally repressed HBV is associated with the B-com-

partment regions of the host genome [10]. Viruses not

only exist episomally, but can also integrate into the

host genome, establishing new cis-contacts. This can

result in activation of the previously silenced genes

[11], or alterations of the native 3D-organization of the

genome– either by integration into the existing archi-

tecture of protein binding sites or by inserting new

ones. Subsequently, these changes can lead to dysregu-

lation of gene activity and cancer development [9, 12].

Another noteworthy aspect of the spatial contacts

of exogenous DNA involves their association with pro-

gressive multifocal leukoencephalopathy (PML) nu-

clear bodies. These bodies facilitate viral replication

and transcription, a phenomenon demonstrated for

various viruses including herpes simplex virus, Simian

Virus 40, adenoviruses, and human papillomavirus

[13, 14]. The study [15] highlighted that the chaper-

one HIRA depositing H3.3 promotes repression of the

naked foreign DNA such as purified plasmid and viral

DNA. This repression is facilitated by the DNA bind-

ing with the HIRA histone chaperone complex into the

PML-bodies within the host nuclei. Concurrently, other

research indicates formation of the viral replication

compartments adjacent to PML, where replication and

transcription processes are executed [16, 17]. Collec-

tively, the existing body of data underscores functional

importance of the three-dimensional contacts between

the exogenous DNA and the genome. However, biology

of extrachromosomal DNA within the nucleus, espe-

cially concerning plasmids, continues to be an area of

further exploration.

Information regarding the biology of extrachro-

mosomal DNA, particularly its interaction with pro-

teins and assembly of the virus-associated compart-

ments, is primarily gathered through immunostaining

and microscopy. However, this approach is not without

its constraints. Resolution offered by these techniques

is often inadequate for smaller structures, including

certain viral genomes and plasmids which are no more

than a few hundred thousand base pairs in size. An al-

ternative method for studying chromatin interactions

employs chromatin conformation capture (3C) technol-

ogies [18]. These technologies offer better resolution,

reaching down to hundreds or thousands of base pairs

for Hi-C, and to the scale of hundreds of base pairs for

circular chromatin conformation capture (4C). Studies

using these methods provide a more detailed view of

the spatial organization of DNA.

Here we introduce a novel method to explore lo-

calization patterns of plasmid DNA, and potentially

other exogenous DNA. This approach is based on the

4C-experiment, utilizing a plasmid as a target vector.

Notable advantage of employing the 4C-methodology

with a plasmid lies in its capability, as we suggest, to

perform further screening experiments – to deliver

plasmids with arbitrary insertions of DNA sequenc-

es into the cell (active and inactive genome regions,

CG-rich and CG-poor regions, Polycomb group protein

binding sites, etc.). We can then observe influence of

these specific nucleotide sequences on the distribution

of plasmids across compartments into the nucleus.

Additionally, this experimental design, involving

delivery of an exogenous insert as a component of a

plasmid into a cell, effectively negates the cis-influence

originating from the adjacent genome regions with di-

verse epigenetic statuses. Thus, it offers a more isolat-

ed and accurate assessment of the impacts attributable

directly to the inserted sequences.

MATERIALS AND METHODS

Plasmid vectors. Modified pUC19 vector was en-

gineered from the original pUC19 by altering two bases

within two NlaIII restriction sites using PCR of plasmid

fragments with primers containing target substitution,

followed by the Gibson Assembly of PCR products into

a circular molecule. The p4CSCS vector was synthe-

sized at the Cloning Facility as two sequences (991bp

and 975bp) that were flanked by homologous regions.

These sequences were joined into a circular molecule

via Gibson Assembly. Maps and sequences of the plas-

mids are provided in the Supplementary materials

(Figs.S1-S3, Online Resource1).

Plasmids were propagated in TOP10 E. coli cells.

We observed that propagation of the p4CSCS vector

necessitated reduction in chloramphenicol concentra-

tion to, specifically, 0.02 µg of chloramphenicol per

1ml of the medium.

Human cell culture and transfection. Exper-

iments were conducted using HEK293T cells, which

were cultured in a DMEM medium supplemented with

TOWARDS DEVELOPMENT OF 4C-BASED METHOD 655

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

10% fetal bovine serum and 1% Pen Strep (all from

Thermo Fisher Scientific, USA). The cells were main-

tained at 37°C and 5% CO

One day prior to transfection, the cells were pas-

saged into a 75 cm

cell culture flask to achieve approx-

imately 70% confluency. Transfection was carried out

in a serum-free DMEM medium devoid of antibiotics.

For transfection mix, we combined 2.4 µl Opti-MEM

(Thermo Fisher Scientific), 18 µg of plasmid DNA, and

36 µg of PEI reagent (2 µg/µl). These conditions were

found to allow efficient transfection of HEK293T cells

(>50% of transfected cells) in control experiments with

a GFP-encoding plasmid (eGFP-n1) co-transfected with

the target vector.

4C-protocol. 24 hours after transfection, cells were

harvested using 0.05% trypsin-EDTA and fixed by 1%

formaldehyde according to the previously published

protocol [19]. Cells were counted and aliquots of 3 mil-

lion cells were snap frozen. We performed key steps of

the experiment: cell lysis, chromatin digestion, chro-

matin ligation, reversal of crosslinking, and DNA puri-

fication (2.2.1-2.2.4) according to the protocol [19] with

minor modifications:

• for chromatin digestion, 50 units of NlaIII was

used.

• biotinylation and “dangling end removal” steps

were skipped.

Digestion of purified DNA. We used slightly dif-

ferent protocols depending on whether p4CSCS-based

vectors (require digestion with MseI) or pUC19-based

vectors (require digestion with TaiI) were transfected

into the cells.

1) Digestion in the case of using p4CSCS-based vec-

tors. For purified DNA digestion, the following com-

ponents were mixed on ice: 5μl 10× CutSmart Buffer,

30 units of MseI, 2μg DNA, ddH

O to 50 μl. The mix-

ture was incubated at 37°C with shaking overnight.

After MseI inactivation by incubation at 65°C 20min,

digested DNA was purified using KAPA Pure Beads (1×)

according to the manufacturer recommendation and

added to the ligation reaction (step “Ligation of puri-

fied DNA”).

2) Digestion in the case of using pUC19-based vec-

tors. In 4C-experiments with modified pUC19 we per-

formed digestion of purified DNA by TaiI: a sample

with 250 ng DNA was supplemented with 1 μl of 10×

restriction buffer 5 units of TaiI, and ddH

O to 10 μl.

The mixture was incubated at 37°C with continuous

shaking overnight. Digestion was terminated by heat

inactivation of the restriction enzyme at 65°C for

20 min. The entire volume of the reaction mixture

(10μl) was added to the ligation reaction. Ligation of

the digested DNA was performed as described in “Li-

gation of purified DNA”.

Ligation of purified DNA. For ligation of purified

DNA the following components were mixed on ice:

10 μl 10% Triton X-100, 10 μl of 10× T4 DNA ligase re-

action buffer, 10 μl 25% PEg, and 1 μl of 10 mg/ml bo-

vine serum albumin, 1 μl 100 mM ATP, 800 units of T4

ligase, 0.5-2 μg DNA, ddH

O to 100 μl (final DNA con-

centration was about 5-20 ng/μl). The mixture was in-

cubated at 16°C overnight, then DNA products were

purified using KAPA Pure Beads (0.8×).

(optional) Additional digestion with DrdI. In our first

rounds of experiments, which were all performed with

pUC19-based vectors and TaiI digestion, we assume

that the fraction of non-informative ligation products

could be reduced by DrdI treatment. Although later

we found that this assumption was not correct, we in-

dicate here that after ligation, products were treated

with DrdI under the following conditions: 100 μl of

ddH

O and 12 units of DrdI endonuclease were added

to the ligase mixture followed by incubation at 37°C

overnight; next DrdI was inactivated by incubation at

80°C for 20 min and DNA products were purified using

KAPA Pure Beads (0.8×).

This step was omitted when using p4CSCS-based

vectors.

Preparation of 4C-libraries for NGS. PCR with prim-

ers containing Illumina adapter sequences was per-

formed for 15-20 cycles. Sequences of the primer are

provided in the Supplementary materials (Online Re-

source 1). PCR products were purified using KAPA Pure

Beads (0.8×). Distribution of DNA-fragments in the ob-

tained 4C-libraries was evaluated by visualization of

the gel after electrophoresis in 1% agarose.

DNA concentration was measured with a Qubit flu-

orimeter. For experiments with modified pUC19 plas-

mid we prepared 4C-libraries for NGS by nested PCR:

we performed the first PCR round with i_F1, i_R1 prim-

ers (see Online Resource 1), then PCR products were

purified using KAPA Pure Beads (0.8×) and one third

of the amount of PCR products served as a DNA matrix

for the second PCR round with P5 and P7 primers for

Illumina sequencing, see Online Resource1).

Experiments with control chromatin and con-

trol plasmids. HEK293T cells were transfected by a

modified pUC19 vector using PEI and fixed 24 h post-

transfection, as described above. Fixed Mus musculus

fibroblasts in approximately equal amounts were in-

troduced to the fixed, transfected HEK293T cells and

co-lysis was conducted as described in the 4C-protocol

above. 200 ng of the control plasmid (modified pUC19

with a cytomegalovirus (CMV) promoter) was added

to the sample before NlaIII digestion. Next, we per-

formed 4C-experiment as described above for all these

components together.

4C-libraries with Klenow treatment. To distin-

guish products formed due to incomplete digestion by

NlaIII, and digested but self-ligated products, we in-

cluded Klenow treatment step into the 4C-protocol de-

scribed above. For this purpose, after NlaIII inactivation,

YAN et al.656

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

centrifugation, and supernatant removal, the follow-

ing components were mixed on ice: 20 μl ×10 NEBuf-

fer 3.1, 1.5× μl of each 10mM dNTPs (dATP, dTTP, dCTP,

dGTP), 50 units of Klenow, ddH

O to 200μl.

The mixture was added to the precipitate, resus-

pended, and incubated at 16°C for 4 h. Then the exper-

iment was continued from the stage of “Blunt end liga-

tion” of the Hi-C 2.0 protocol [19] up to and including

the DNA purification stage and further according to

the above described 4C-protocol with plasmid.

4C-data processing and analysis. Raw reads

were processed using cutadapt and then aligned to a

reference that included the human hg38 genome and

plasmid vector sequences (either modified pUC19 or

p4CSCS). For experiments involving control chromatin,

the reference also incorporated the mouse mm10 ge-

nome, modified pUC19, and modified pUC19 with the

CMV-promoter sequence. Alignment was performed

using the BWA mem. Alignment statistics were gath-

ered with samtools idxstats. To visualize the alignment

results, we employed IGV2.16.1 software.

In our protocol, two ligation events should occur:

first ligation at the NlaIII site at the fixed chromatin

condition and second ligation at the MseI or TaiI site

during the circularization step. Products of the first li-

gation reaction represent spatial colocalization of DNA

fragments in the nucleus, whereas the second ligation

can either occur as intramolecular ligation producing

circular DNA or intermolecular ligation between ran-

dom DNA ends. Thus, products of the second ligation

do not add information about spatial contacts of the

plasmid, but may reflect random collisions resulting

from diffusion processes. Therefore, we discarded se-

quences ligated to the MseI or TaiI site, and only used

sequences ligated to the NlaIII site. According to the

plasmid vector design, these sequences always start

from the p7-adapter (Figs.S1-S3, Online Resource1).

RESULTS

Design of 4C-experiment with plasmid. Weaimed

to develop an approach based on using a 4C-method

to identify contacts between the plasmid DNA and the

host genome. Design of the experiment is shown in

Fig. 1a. Initially, cells are transfected with a plasmid

vector. We anticipate that within the nuclei, plasmids

are secured by host cellular proteins, and biophysical

characteristics of these proteins dictate localization of

the plasmid (I). Subsequently, spatial contacts within

chromatin, including interactions between the plasmid

and DNA, are fixed with formaldehyde (II). According

to the 4C-protocol (detailed in methods), sequential di-

gestion and proximity ligation reactions are conducted

on the fixed chromatin (III-IV). After this, the cross-

links are reversed; the DNA is digested with a second

restriction enzyme and ligated in solution (V). This

results in formation of circular molecules, containing

a fragment of the plasmid DNA and a segment of the

genome with which the plasmid was in contact (VI).

By amplification and sequencing of these circular tem-

plates using NGS technology, we expect to detect inter-

actions between the plasmid DNA and the genome.

According to the described design, the vector used

in the 4C-experiment should include two distinct re-

striction sites: the first is hydrolyzed during the initial

DNA digestion in the fixed chromatin (E1, Fig.1a), and

the second during the subsequent digestion of the plas-

mid DNA following cross-links reversal (E2, Fig. 1a).

A crucial aspect to consider is placement of the PCR

primer annealing region, which should be located be-

tween these two restriction sites, near to their posi-

tions. Furthermore, there must be neither E1 nor E2

restriction sites between the primer annealing regions.

Moreover, both E1 and E2 should be frequent cutter

sites, i.e.,4-base cutters, to allow high resolution of the

4C-analysis (determined by E1 site frequency), and ef-

ficient amplification of 4C-fragments (determined by

E2 site frequency). Finally, the E1-specific enzyme has

to be able to digest formaldehyde-fixed chromatin.

To validate the proposed design, we performed

4C-experiments with the pUC19 plasmid vector.

Weused NlaIII and TaiI restriction sites as E1 and E2

target sites of the experiment, respectively. In addition

to the two designated NlaIII restriction sites, the pUC19

vector contains additional NlaIII and TaiI sites (nine

and three, respectively), including those present in the

region for PCR primer annealing (Fig.1b). Weexpect-

ed that additional NlaIII and TaiI sites outside the tar-

get region (where binding sites of the primers used for

library amplification are located) would not interfere

with the 4C-experiment, as opposed to the sites locat-

ed between the primers. Therefore, we engineered a

modified vector based on pUC19, wherein the NlaIII

(-CATG-) restriction sites located between the primers

were removed (Fig.1b).

Estimating background noise level in 4C-pro-

tocol with plasmid. The 4C-experiment design with

plasmid assumes that restriction and ligation stages

occur under conditions of fixed chromatin, therefore

the resulting products reflect spatial co-localization

of the plasmid DNA and the host genome inside the

nucleus. However, it is also possible that part of the

contacts is formed in solution as a result of diffusion

processes after chromatin fixation, causing random

collisions and ligations of the plasmid and the genom-

ic DNA fragments. These contacts do not reflect bio-

logical preferences of the plasmid distribution in the

nucleus.

To estimate the scale of such ligation events, we

performed a control experiment, shown in Fig. 2a.

Human HEK293T cells, transfected with the modified

TOWARDS DEVELOPMENT OF 4C-BASED METHOD 657

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

Fig. 1. Design of 4C-experiment with plasmid vector. a)[I-VI]overall experiment design. b)Restriction sites scheme of the plas-

mid vectors generated in this work. Plasmid maps and sequences are provided in the Supplementary materials (Figs.S1-S3,

Supplementary text 2, see Online Resource1).

pUC19 vector, underwent chromatin fixation 24 h post-

transfection. Next, the control chromatin (fixed chro-

matin of a non-human species) and the control plas-

mid (plasmid with the sequence different from pUC19

vector) were added to these fixed HEK293T samples.

Finally, we prepared 4C-libraries as described in the

experiment design section above (Fig. 1a). Given that

the control chromatin of non-human species was add-

ed after fixation, any ligations of the pUC19 vector

(which was transfected pre-fixation) with non-human

chromatin represent random collision events. Simi-

larly, ligations of the control plasmid (added post-fixa-

tion) with the HEK293T chromatin and with chromatin

of a non-human species are due to random collisions

as well.

Therefore, the contacts of plasmid no. 1 (modi-

fied pUC19 transfected into the nucleus) reflect both

spatial localization of the plasmid in living cells (by

ligation with human chromatin) and possible random

diffusion processes (by ligation with chromatin of hu-

man and non-human species). The contacts of plasmid

no. 2 added to the cells after fixation with chromatin

are not conditioned by biological regularities and are

only determined by random diffusion processes (by li-

gation with human chromatin and with chromatin of

anon-human species).

We compared how often the plasmid no. 1, trans-

fected into human cells, and plasmid no. 2, added to the

cells after fixation, contact with the human and non-

human chromatin. Our results, presented in Fig. 2b,

demonstrate that 80% of the plasmid no. 1 contacts

are with the human chromatin, whereas the plasmid

no. 2 interacts with the human chromatin only in 64%

of cases (percentages are average for three replicas;

YAN et al.658

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

Fig. 2. Estimation of “random” contacts fraction. a) Design of the experiment: plasmid no. 1 was transfected into HEK293T

cells(I), then spatial contacts of DNA and proteins in the nucleus were fixed with formaldehyde(II). Fixed Mus musculus fibro-

blasts(III) and plasmid No. 2(IV) were added to the obtained sample, and the 4C-experiment was performed. b)Plasmid no. 1

and plasmid no. 2 differ in the ratio of the number of contacts with H. sapiens and M. musculus chromatin The diagrams show

percentage of the plasmid contacts with the corresponding chromatin, averaged over three independent experiments. Results

for the individual replicates are presented in Fig.S4 in the Online Resource1. Total number of contacts is taken as 100%.

for all three replicas difference of contact frequencies

was significant according to Fisher’s exact test, p-value

is negligible; see Supplementary notes for details, On-

line Resource1). These observations suggest that part

of the contacts between the plasmid transfected into

cells and the chromatin detected in the experiment is

due to the spatial proximity in the nucleus, i.e., cannot

be explained by random collisions of the plasmid DNA

with DNA in solution. We also quantify the fraction of

random collisions (see Supplementary notes, Online

Resource1).

4C-vector optimization. Analyzing the data ob-

tained in the 4C-experiment with the modified pUC19

we faced the problem that the majority (about 90%)

of the reads did not contain the human genome se-

quence. Instead, the reads aligned to the specific re-

gions of the plasmid. This issue was not only observed

when using control chromatin and control plasmid,

but also in the experiments involving HEK293T cells

transfected with the modified pUC19 vector and sub-

jected to the 4C-protocol without controls. Detailed

analysis of the reads distribution and structure re-

vealed that this phenomenon could be traced back to

the ligation of various plasmid fragments during the

in-chromatin ligation stage. It occurs because the plas-

mid DNA is not only hydrolyzed at the intended NlaIII

and TaiI restriction sites, but also at other NlaIII and

TaiI sites present in the plasmid. Under conditions pro-

moting proximity ligation, fragments originating from

a single plasmid molecule are more likely to undergo

ligation in cis, a phenomenon we have named “self-

ligation.” Consistent with the significant prevalence of

cis-interactions over the trans-contacts, which is typi-

cal for the 3C-data [20], reads containing only plasmid

sequences make up to about 90% of the total data.

High frequency of ligations occurring between

the plasmid DNA fragments, as described above, can

be attributed to the predominance of cis (within the

plasmid) over trans (plasmid-genome) spatial contacts

and therefore is unavoidable in the 3C-experiment.

TOWARDS DEVELOPMENT OF 4C-BASED METHOD 659

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

However, this problem could be overcome if the plas-

mid vector lacks multiple digestion sites. In such cases,

cis-ligation would restore the original plasmid mole-

cule, which would then be excluded during the PCR

and NGS steps due to its length (>1000bp). Therefore,

to solve the problem of plasmid self-ligation, we aimed

to redesign the plasmid sequence to contain only sin-

gle NlaIII and MseI restriction sites (in this vector de-

sign, we use MseI instead of TaiI at the purified DNA

digestion step). This task is challenging because both

enzymes are frequent cutters (which is essential for

the resolution and detectability of contacts, see above).

Consequently, multiple cut sites are present in all func-

tional elements of the available plasmid vectors.

To overcome this limitation, we substituted the

ampicillin resistance gene with the chloramphenicol

resistance cassette, devoid of restriction sites in the

promoter region. However, the replication origin and

body of the chloramphenicol resistance gene still con-

tained NlaIII and MseI restriction sites. We introduced

single-nucleotide substitutions within the gene body,

ensuring the amino acid sequence remained unal-

tered. Choice of alternative nucleotides was based on

the codon frequencies in E. coli [21]. In the replication

origin, the substitutions were made arbitrarily, given

the absence of the straightforward criterion to predict

their functional implications. As a result, we designed a

sequence of a new plasmid adapted for 4C-experiment,

containing only two target restriction sites, p4CSCS:

(plasmid for 4C with Single Cutter Sites), depicted in

Fig. 1b (plasmid map and sequence see in Fig. S3 and

in supplementary text 2, Online Resource 1). It con-

tains primer binding regions and essential elements

for reproduction in E. coli, such as replication origin

and chloramphenicol resistance gene. Total length of

the vector is 1826bp.

We used this improved plasmid vector in the

4C-experiment with HEK293T cells. The obtained data

contained a substantial proportion of contacts between

the plasmid and the genome, approximately 81%.

Thus, we confirm that the observed in previous exper-

iments high frequency of reads mapped to the plasmid

was due to the presence of multiple restriction sites in

the vector. Obtained results clearly show the advan-

tage of the redesigned plasmid sequence for 4C-exper-

iments.

Although contacts between the plasmid and the

genome are well represented, reads including exclu-

sively plasmid sequences persist, accounting for about

20% of the data. According to visualization in IGV,

the majority of these reads are mapped upstream of

the NlaIII restriction site. This observation raises the

question of the origin of these reads: whether they are

formed due to incomplete digestion by NlaIII, or be-

cause the plasmid molecules were hydrolyzed but then

ligated back to form a whole plasmid molecule again?

To confirm which hypothesis is accurate, we pre-

pared 4C-libraries where chromatin was treated with

the Klenow fragment of E. coli DNA polymerase I after

digestion by the NlaIII enzyme. Hydrolysis with the

NlaIII enzyme produces 3′-ends that are removed by

the 3′-5′ exonuclease activity of the Klenow fragment.

Subsequent ligation of the plasmid ends generates a

specific sequence signature distinct from the undi-

gested plasmid sequence. Therefore, sequencing the

ligation products should distinguish between the mol-

ecules from the undigested plasmid (containing the

sequence 5′…TCTGAC-CATG-AGGAGA…3′ with the orig-

inal NlaIII 5′…-CATG-…3′ restriction site subsequence)

and the molecules that were hydrolyzed by NlaIII but

then ligated back (containing the Klenow ends resec-

tion signature 5′…TCTGAC-AGGAGA…3′, without the

NlaIII restriction site).

Sequencing data from six independent experi-

ments showed that the number of non-hydrolyzed

products (with a mean fraction of 4.82% of all reads

analyzed) was almost 30 times greater than the num-

ber of molecules that were hydrolyzed and ligated

back together (mean fraction of 0.17% of all reads an-

alyzed). This indicates that the reads containing an

intact plasmid sequence are mainly due to inefficient

hydrolysis by the NlaIII enzyme.

DISCUSSION

In this study, we present a method that enables

detection of the contacts between the plasmid DNA

introduced into the cell and the host genome. Our

preliminary experiments show promising results: the

4C-protocol allows efficient enrichment of the con-

tacts between the plasmid and host genome DNA. Al-

though interactions representing contacts between the

genomic loci are about a million times more frequent

than between the genome and plasmid (according to

the ratio of human genome length to the approximate

length of both modified pUC19 and p4CSCS plasmid), a

large portion of the contacts detected in the 4C-exper-

iment contains plasmid sequences. Moreover, optimal

design of the plasmid vector allows to enrich the plas-

mid-genome interactions over the plasmid-plasmid in-

teractions, though the latter occur in cis and therefore

is expected to have higher frequency. Looking ahead,

precision and informational yield of this method could

be augmented through refinement of the chromatin

digestion process. This assumption is supported by our

observation that the fraction of non-informative prod-

ucts originates from the undigested plasmid.

In addition, it should be noted that the noise in

3C-data is expected to be uniformly distributed and,

therefore, should not hinder detection of the increased

frequency of contacts between the plasmid and specific

YAN et al.660

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

genomic loci. However, high noise level may require

deeper sequencing to detect specific interactions.

We envision several applications of the developed

method.

• First, the proposed method can be used to in-

vestigate localization of the plasmid molecules them-

selves. Despite the widespread usage of plasmids as

vectors for exogenous gene expression in genetic engi-

neering and research, distribution of plasmids in the

nucleus remains largely unknown [22-24].

• Second, the proposed experiment design allows

us to test how DNA sequences with different epigen-

etic properties (active and inactive genome regions

such as GC-rich, GC-poor regions, Polycomb, and HP1

binding sites, etc.) embedded in the plasmid will af-

fect plasmid localization. This may be applicable to

the study of compartments formed as a result of phase

separation in the nuclei. Although it is clear that the

active and inactive chromatin segregate in the nucle-

us, it is still an open question how many phases (or

subcompartments) are formed in total [25, 26]. Phase

separation of the nucleus is being actively studied

from the protein perspective. Compartmentalization

can also be considered from the viewpoint of DNA

motifs to which proteins involved in phase separation

are bound. Several DNA motifs were shown to be es-

sential for the loci compartmentalization. For exam-

ple, DNA sequences that define compartmentalization

profile for the Polycomb-repressed loci in Drosophila

embryos were identified using the Hi-C and ChIP-seq

methods and confirmed using genome editing [27]. Be-

sides that, genetic determinants of compartmentaliza-

tion remain unknown. We suggest that the 4C-method

with plasmid could be applied to identify genetic de-

terminants of compartmentalization by testing effect

of different DNA sequences on the pattern of plasmid

contacts with the genome. In perspective, the method

could be used as a basis for large-scale screening stud-

ies and will allow us to compile a profile of the contri-

bution of each locus to chromatin compartmentaliza-

tion throughout the genome.

• DNA methylation alters three-dimensional or-

ganization of the genome and chromatin accessibility

by recruiting proteins such as HP1, which form het-

erochromatin blocks as a result of phase separation

[25, 28, 29]. Methylation of CTCF binding sites disrupts

its association with DNA, leading to changes in the ge-

nome architecture [30]. Overall, CpG methylation af-

fects the ability of TFs to bind to DNA in both negative

and positive ways in vitro [31]. However, this has not

been confirmed in vivo [28]. We propose integration

of the 4C-method with the in vitro 5mC-methylated

plasmid DNA as a strategic approach for in vivo local-

ization of the methylated DNA loci and clarification

whether they form distinct subcompartments by clus-

tering together.

• The 4C-method with plasmid can be used to

study DNA repair. A plasmid with modified nucleo-

tides, such as 8-oxoguanine inclusions or other mod-

ifications, is likely to be localized in the specific DNA

repair compartments assembly of which is initiated

by the activated PARP1 and FUS [32]. The 4C-experi-

ment can be used to capture the loci attracted to the

DNA repair compartment, as well as dynamics of its

formation.

To summarize, we are optimistic that the refined

4C-experiment with plasmid, as described in this study,

would be instrumental in the domains of 3D-genomics,

transcription regulation, and epigenetics.

Supplementary information. The online version

contains supplementary material available at https://

doi.org/10.1134/S0006297924040059.

Acknowledgments. We acknowledge infrastruc-

ture and resources provided by the collective usage

center of the Institute of Cytology and Genetics, Si-

berian Branch of the Russian Academy of Sciences,

121031800061-7 (Mechanisms of genetic control of

development, physiological processes, and behavior

in animals) for running next generation sequencing

experiments. Computational data analysis was per-

formed using the high-throughput computing nodes

of the Novosibirsk State University (supported by the

Ministry of Science and Higher Education of the Rus-

sian Federation, grant no. FSUS-2024-0018).

Contributions. V.F., P.S., and A.Y. conceived the

study. A.Y. performed experiments with the help from

P.S. and M.G. P.B. developed a computational data anal-

ysis pipeline. A.Y. performed data analysis with the

help from P.B., P.S., and V.F. supervised the study and

analyzed the data. A.Y., P.S., and V.F. wrote the manu-

script. All authors edited the manuscript and approved

its final version.

Funding. This work was supported by the Russian

Science Foundation (project no.22-14-00242).

Ethics declarations. This work does not con-

tain any studies involving human and animal sub-

jects. Theauthors of this work declare that they have

noconflicts of interest.

REFERENCES

1. Kabirova,E., Nurislamov,A., Shadskiy,A., Smirnov,A.,

Popov, A., etal. (2023) Function and evolution of the

loop extrusion machinery in animals, Int.J. Mol. Sci.,

24, 5017, doi:10.3390/ijms24055017.

2. Nuebler,J., Fudenberg,G., Imakaev,M., Abdennur,N.,

and Mirny, L.A. (2018) Chromatin organization by an

interplay of loop extrusion and compartmental seg-

regation, Proc. Natl. Acad. Sci. USA, 115, E6697-E6706,

doi:10.1073/pnas.1717730115.

TOWARDS DEVELOPMENT OF 4C-BASED METHOD 661

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

3. Fishman, V., Battulin, N., Nuriddinov, M., Maslo-

va, A., Zlotina, A., et al. (2019) 3D organization of

chicken genome demonstrates evolutionary con-

servation of topologically associated domains and

highlights unique architecture of erythrocytes’ chro-

matin, Nucleic Acids Res., 47, 648-665, doi: 10.1093/

nar/gky1103.

4. Ryzhkova,A., Taskina,A., Khabarova,A., Fishman,V.,

and Battulin,N. (2021) Erythrocytes 3D genome orga-

nization in vertebrates, Sci. Rep., 11, 4414, doi:10.1038/

s41598-021-83903-9.

5. Razin, S.V., and Gavrilov, A.A. (2020) The role of liq-

uid-liquid phase separation in the compartmental-

ization of cell nucleus and spatial genome organiza-

tion, Biochemistry (Moscow), 85, 643-650, doi:10.1134/

S0006297920060012.

6. Kantidze, O.L., and Razin, S. V. (2020) Weak interac-

tions in higher-order chromatin organization, Nucleic

Acids Res., 48, 4614-4626, doi:10.1093/nar/gkaa261.

7. Nuriddinov, M., and Fishman, V. (2019) C-InterSec-

ture-a computational tool for interspecies compari-

son of genome architecture, Bioinformatics (Oxford,

England), 35, 4912-4921, doi: 10.1093/bioinformatics/

btz415.

8. Lukyanchikova, V., Nuriddinov, M., Belokopytova, P.,

Taskina, A., Liang,J., et al. (2022) Anopheles mosqui-

toes reveal new principles of 3D genome organiza-

tion in insects, Nat. Commun., 13, 1960, doi: 10.1038/

s41467-022-29599-5.

9. Dias, J.D., Sarica,N., Cournac,A., Koszul,R., and Neu-

veut,C. (2022) Crosstalk between hepatitisB virus and

the 3D genome structure, Viruses, 14, 445, doi:10.3390/

v14020445.

10. Tang,D., Zhao,H., Wu,Y., Peng,B., Gao,Z., etal. (2021)

Transcriptionally inactive hepatitis B virus episome

DNA preferentially resides in the vicinity of chromo-

some 19 in 3D host genome upon infection, Cell Rep.,

35, 109288, doi:10.1016/j.celrep.2021.109288.

11. Sokol, M., Wabl, M., Ruiz, I. R., and Pedersen, F. S.

(2014) Novel principles of gamma-retroviral inser-

tional transcription activation in murine leukemia

virus-induced end-stage tumors, Retrovirology, 11, 36,

doi:10.1186/1742-4690-11-36.

12. Razin, S.V., Gavrilov, A.A., and Iarovaia, O.V. (2020)

Modification of nuclear compartments and the 3D ge-

nome in the course of a viral infection, Acta Naturae,

12, 34-46, doi:10.32607/actanaturae.11041.

13. Everett, R. D. (2013) The spatial organization of

DNA virus genomes in the nucleus, PLoS Pathog., 9,

e1003386, doi:10.1371/journal.ppat.1003386.

14. Corpet, A., Kleijwegt, C., Roubille, S., Juillard, F., Jac-

quet,K., etal. (2020) PML nuclear bodies and chroma-

tin dynamics: catch me if you can!, Nucleic Acids Res.,

48, 11890-11912, doi:10.1093/nar/gkaa828.

15. Rai, T.S., Glass,M., Cole, J.J., Rather, M.I., Marsden,M.,

etal. (2017) Histone chaperone HIRA deposits histone

H3.3 onto foreign viral DNA and contributes to anti-vi-

ral intrinsic immunity, Nucleic Acids Res., 45, 11673-

11683, doi:10.1093/nar/gkx771.

16. Schmid, M., Speiseder, T., Dobner, T., and Gonzalez,

R.A. (2014) DNA virus replication compartments,

J.Vi-

rol., 88, 1404-1420, doi:10.1128/JVI.02046-13.

17. Charman,M., and Weitzman, M.D. (2020) Replication

compartments of DNA viruses in the nucleus: loca-

tion, location, location, Viruses, 12, 151, doi:10.3390/

v12020151.

18. Kempfer,R., and Pombo, A. (2020) Methods for map-

ping 3D chromosome architecture, Nat. Rev. Genet., 21,

207-226, doi:10.1038/s41576-019-0195-2.

19. Belaghzal, H., Dekker, J., and Gibcus, J.H. (2017) Hi-C

2.0: an optimized Hi-C procedure for high-resolution

genome-wide mapping of chromosome conformation,

Methods (San Diego, Calif.), 123, 56-65, doi: 10.1016/

j.ymeth.2017.04.004.

20. Gridina,M., Mozheiko,E., Valeev,E., Nazarenko, L.P.,

Lopatkina, M. E., et al. (2021) A cookbook for DNase

Hi-C, Epigenet. Chromatin, 14, 15, doi:10.1186/s13072-

021-00389-5.

21. Gvritishvili, A.G., Leung, K.W., and Tombran-Tink,J.

(2010) Codon preference optimization increases het-

erologous PEDF expression, PLoS One, 5, e15056,

doi:10.1371/journal.pone.0015056.

22. Prajapati, H. K., Kumar, D., Yang, X.-M., Ma, C.-H.,

Mittal, P., et al. (2020) Hitchhiking on condensed

chromatin promotes plasmid persistence in yeast

without perturbing chromosome function, bioRxiv,

doi:10.1101/2020.06.08.139568.

23. Gracey Maniar, L.E., Maniar, J. M., Chen, Z.-Y., Lu, J.,

Fire, A.Z., etal. (2013) Minicircle DNA vectors achieve

sustained expression reflected by active chroma-

tin and transcriptional level, Mol. Ther., 21, 131-138,

doi:10.1038/mt.2012.244.

24. Dean, D.A. (1997) Import of plasmid DNA into the nu-

cleus is sequence specific, Exp. Cell Res., 230, 293-302,

doi:10.1006/excr.1996.3427.

25. Mladenova,V., Mladenov,E., and Russev,G. (2009) Or-

ganization of plasmid DNA into nucleosome-like struc-

tures after transfection in eukaryotic cells, Biotech-

nol. Biotechnolog. Equip., 23, 1044-1047, doi: 10.1080/

13102818.2009.10817609.

26. Hildebrand, E. M., and Dekker, J. (2020) Mechanisms

and functions of chromosome compartmentaliza-

tion, Trends Biochem. Sci., 45, 385-396, doi: 10.1016/

j.tibs.2020.01.002.

27. Erdel, F., and Rippe, K. (2018) Formation of chroma-

tin subcompartments by phase separation, Biophys.J.,

114, 2262-2270, doi:10.1016/j.bpj.2018.03.011.

28. Ogiyama,Y., Schuettengruber,B., Papadopoulos, G.L.,

Chang, J.-M., and Cavalli, G. (2018) Polycomb-depen-

dent chromatin looping contributes to gene silencing

during Drosophila development, Mol. Cell, 71, 73-88.

e5, doi:10.1016/j.molcel.2018.05.032.

YAN et al.662

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

29. Mattei, A. L., Bailly, N., and Meissner, A. (2022) DNA

methylation: a historical perspective, Trends Genet.,

38, 676-707, doi:10.1016/j.tig.2022.03.010.

30. Rountree, M. R., and Selker, E. U. (2010) DNA methyl-

ation and the formation of heterochromatin in Neu-

rospora crassa, Heredity, 105, 38-44, doi: 10.1038/

hdy.2010.44.

31. Phillips, J. E., and Corces, V. G. (2009) CTCF: mas-

ter weaver of the genome, Cell, 137, 1194-1211,

doi:10.1016/j.cell.2009.06.001.

32. Singatulina, A. S., Hamon, L., Sukhanova, M.V., Des-

forges,B., Joshi,V., Bouhss,A., Lavrik, O.I., and Pas-

tré, D. (2019) PARP-1 activation directs FUS to DNA

damage sites to form PARG- reversible compartments

enriched in damaged DNA, Cell Rep., 27, 1809-1821,

doi:10.1016/j.celrep.2019.04.031.

Publisher’s Note. Pleiades Publishing remains

neutral with regard to jurisdictional claims in pub-

lished maps and institutional affiliations.