Article

ISSN 0006-2979, Biochemistry (Moscow), 2024, Vol. 89, No. 4, pp. 585-600 © Pleiades Publishing, Ltd., 2024.

585

REVIEW

Cohesin Complex:

Structure and Principles of Interaction with DNA

Arkadiy K. Golov

1,2,a

* and Alexey A. Gavrilov

1,b

Institute of Gene Biology, Russian Academy of Sciences, 119334 Moscow, Russia

Technion – Israel Institute of Technology, 3525433 Haifa, Israel

e-mail: golovstein@gmail.com

e-mail: aleksey.a.gavrilov@gmail.com

Received October 14, 2023

Revised February 19, 2024

Accepted February 23, 2024

Abstract— Accurate duplication and separation of long linear genomic DNA molecules is associated with a number

of purely mechanical problems. SMC complexes are key components of the cellular machinery that ensures decat-

enation of sister chromosomes and compaction of genomic DNA during division. Cohesin, one of the essential eu-

karyotic SMC complexes, has a typical ring structure with intersubunit pore through which DNA molecules can be

threaded. Capacity of cohesin for such topological entrapment of DNA is crucial for the phenomenon of post-rep-

licative association of sister chromatids better known as cohesion. Recently, it became apparent that cohesin and

other SMC complexes are, in fact, motor proteins with a very peculiar movement pattern leading to formation of

DNA loops. This specific process has been called loop extrusion. Extrusion underlies multiple functions of cohesin

beyond cohesion, but molecular mechanism of the process remains a mystery. In this review, we summarized the

data on molecular architecture of cohesin, effect of ATP hydrolysis cycle on this architecture, and known modes of

cohesin–DNA interactions. Many of the seemingly disparate facts presented here will probably be incorporated in

a unified mechanistic model of loop extrusion in the not-so-distant future.

DOI: 10.1134/S0006297924040011

Keywords: SMC complexes, cohesin, SMC subunits, kleisin, HAWK subunits, cohesion, topological entrapment, loop

extrusion, DNA gripping state

Abbreviations: HAWK,HEAT protein associated with Kleisin;

SMC,structural maintenance of chromosomes proteins.

* To whom correspondence should be addressed.

INTRODUCTION

Cohesin is a paneukaryotic protein complex be-

longing to the group of structural maintenance of

chromosomes proteins (SMC) participating in the pro-

cesses of individualization of sister chromatids, chro-

mosome decatenation, mitotic and meiotic cohesion,

establishment and maintenance of specific DNA fold-

ing in cells, repair of double-strand breaks, and sta-

bilization of stalled replication forks [1,2]. Similar to

other SMC complexes, cohesin, on the one hand, com-

prises a multisubunit ABC-ATPase, and, on the other

hand,– a non-sequence-specific DNA-binding complex.

Binding and hydrolysis of ATP are coupled with con-

formational changes in cohesin, as well as with changes

in affinity of different components of the complex to

DNA. The repeating binding/ATP hydrolysis cycles

result in the directional movements of the complex

along the DNA thread; hence, cohesin is a motor pro-

tein– DNA translocase.

Cohesin performs two different functions in the

eukaryotic cells: mediating sister chromatid cohe-

sion and so-called chromatin loop extrusion (Fig. 1a).

The process of extrusion starts with the capture of a

small DNA loop followed by its processive enlargement

by pulling DNA flanking regions inside [3-6]. An active

SMC complex is located at the basis of growing loops,

and their size could reach up to hundreds of thou-

sands of base pairs (Fig. 1a (2)). Extrusion is coupled

with ATP hydrolysis and represents a universal bio-

chemical activity of all SMC complexes[1]. Most likely,

extrusion appeared at the dawn of cellular evolution

as a way of post-replicative separation of the massive

molecules of genomic DNA(Fig.1b); in this process the

GOLOV, GAVRILOV586

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

Fig. 1. Sister chromatid cohesion and DNA loop extrusion is mediated by SMC complexes. a)Two basic activities of cohesin:

cohesion(1) and extrusion (2). b)Extrusion mediated by SMC complexes, and activity of type II DNA topoisomerases ensure

post-replicative individualization of sister genomes in all cells, prokaryotic and eukaryotic.

SMC complexes cooperate with the typeII topoisomer-

ases. Cohesion of sister chromatids mediated by cohes-

in rings (Fig. 1a (1)), most likely, does not depend on

the extrusion activity of the complex [7, 8], although

ATP hydrolysis is also required for the loading of co-

hesin rings onto DNA [7,9].

Information on the protein structure is important

for understanding their activity and even more im-

portant for understanding activity of motor proteins.

In this review the data on cohesin structure as well as

current understanding of its interactions with DNA,

conformational changes, and how interactions with

DNA and binding/ hydrolysis of ATP control these con-

formational changes are presented. Detailed data on

the biological role of the cohesin-dependent extrusion,

as well as molecular mechanism of this process are

presented in the second part of this review published

in the same issue of this journal[10].

ANNULAR CORE COHESIN TRIMER COMPLEX

All eukaryotic genomes encode at least four SMC

complexes from three different classes: two variants

of cohesin (mitotic and meiotic), condensin I, and

SMC5/6 complex[3,4] (table). Trimer composed of two

SMC subunits and one kleisin subunit forms an invari-

ant basis of SMC complexes [2, 3,11]. SMC proteins[12]

and kleisins[13] represent two unrelated protein fam-

ilies. Different classes of SMC complexes are separat-

ed by hundreds of millions of years of evolution and

lost their homology in the larger part of their amino

acid sequences, while similarity within the short func-

tionally important motifs was retained, as well as their

common basic structure [14]. Nevertheless, a signifi-

cant degree of conservation has been observed within

the six main subfamilies of eukaryotic SMC proteins

(SMC1-6) and three subfamilies of eukaryotic kleisins

(kleisinsα-γ). The especially high degree of conserva-

tion is observed within the SMC1- and SMC3-subfam-

ilies. For example, the human and mouse SMC3 are

practically identical.

In the case of cohesin (here and further we con-

sider the better described mitotic variant of the com-

plex) the pair of SMC proteins is represented by the

SMC1 (Smc1)–SMC3 (Smc3)

heterodimer and the klei-

sin subunit– by the RAD21 protein (Scc1). SMC proteins

are rod-like molecules with length of about 50nm with

two globular domains at the ends connected via rela-

tively labile fibrillar structure (Fig.2a)[16-18]. Within

the rod-like structure of the SMC protein 1200-1300aa

long polypeptide chain is folded back on itself so that

Names of human proteins are used in the main text and names of homologous Saccharomyces cerevisiae proteins are given

in parenthesis (at first mention).

COHESIN COMPLEX STRUCTURE 587

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

Subunit composition of paneukaryotic SMC complexes

Complex →

Meiotic cohesin Mitotic cohesin CondensinI SMC5/6 complex

↓ Subunit

ν-SMC SMC3 (Smc3) SMC3 (Smc3) SMC2 (Smc2) SMC6 (Smc6)

κ-SMC SMC1 (Smc1) SMC1α (Smc1) SMC4 (Smc4) SMC5 (Smc5)

Kleisin REC8 (Rec8) RAD21 (Scc1) CAP-H (Brn1) NSE4A/B (Nse4)

HAWK

PDS5A/B (Pds5) NIPBL (Scc2) and PDS5A/B (Pds5) CAP-D2 (Ycs4) –

HAWK

STAG3 (Scc3) STAG1/2 (Scc3) CAP-G (Ycg1) –

KITE

– – – NSE5 (Nse5)

KITE

– – – NSE6 (Nse6)

Note. HAWK-subunits are present only in cohesins and condensins, while prokaryotic SMC complexes and SMC5/6 complex have

auxiliary subunits in their composition belonging to the group of KITE-proteins[15]. Names of the human subunits are present-

ed as well as names of homologous subunits of Saccharomyces cerevisiae (in parenthesis).

Fig. 2. Subunit structure of cohesin complex. a) Folding and main structural features of SMC proteins. b)General structure

of three-part cohesin ring and interaction of HAWK-subunits with it. c) General structure of hook-shaped HAWK-subunits

inSMCcomplexes.

one of the terminal globular domains is formed as a

result of interactions of N- and C-ends, and another is

formed by the uninterrupted sequence located approx-

imately in the middle of the linear sequence [17, 19].

The first of the globular domains was termed head

domain, and the second – hinge domain; polypeptide

chain of the SMC protein makes a 180° turn within the

latter. Fibrillar structure connecting two globular do-

mains comprises an intramolecular coiled-coil struc-

ture termed arm domain.

Stable dimerization of the SMC proteins is achieved

via homotypic interactions between the hinge domains

[17, 18]. Coiled-coil arm domains are directed to the

same direction away from the interacting hinge do-

mains; hence, the SMC1–SMC3 dimer in the absence

of ATP comprises a V-shaped structure with dimerized

GOLOV, GAVRILOV588

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

Fig. 3. Diversity of cohesin conformational states. a)O-, E-, and J-configurations of cohesin head domains and transitions be-

tween them, RAD21-subunit is not shown for better clarity. S-K-ring and subcompartments (E-S, E-K, and J-K) formed as a result

of head domains engagement are shown in pictograms. b)Major conformational states of cohesin detected using microscopy.

hinge domains at one pole and pair of separated head

domains at another pole [17,20].

The RAD21-kleisin subunit has length of approx-

imately 500-700 aa with two structured domains at

N- and C-ends, and extended mostly disordered region

between them [13]. The N-terminal domain of kleisin

interacts with the region connecting the head and arm

domains of SMC3 subunit, while the C-terminal do-

main– with the head domain of SMC1(Fig.2b)[21].

Molecules of the three core subunits of cohesin

form a closed annular structure through stable inter-

actions between the termini[17,22]. Elongated shape

of each of the core subunits results in formation of co-

hesin conformations in which it assumes an expanded

topologically closed intersubunit compartment called

S-K-ring(Fig.3a). Opening of the S-K-ring, also called

S-K-compartment, allows passage of globular parti-

cles with diameter of up to around 10 nm through

it[23,24]. It was shown that one or two DNA threads

can be entrapped within the S-K-compartment, thus

the ring-shaped complex can be put on the DNA as a

bead on a string. Inthe G2-phase of cell cycle such to-

pological interactions of cohesin molecules with the

pairs of sister chromatids mediate cohesion [7].

Hinge domains of the SMC proteins comprise

compact structures consisting of two (N- and C-termi-

nal) subdomains of α/β-class; each of the subdomains

contains a small β-sheet [17, 18]. The N-terminal sub-

domain interacts with the C-terminal subdomain of its

dimerization partner with merger of their β-sheets.

Hence, two separate surfaces are involved in the in-

teraction between two hinge domains of the dimer.

Dueto the fact that the two hinge subdomains are sep-

arated by a small groove, the dimerized hinge domains

look like a toroidal structure with the 2-fold rotation-

al pseudosymmetry (Fig.2b). Small channel inside the

hinge domain dimer contains functionally important

positively charged amino acids, which are assumed to

be able to participate in electrostatic interactions of

the complex with DNA. Asymmetry in the heterodimer

of cohesin SMC subunits allows distinguishing two in-

teraction surfaces of the hinge domains. One of them,

which is at a larger distance from the head domains

is commonly known as a northern one, and another–

asa southern (Fig.2b).

The arm domain of each SMC protein compris-

es a pair of antiparallel coiled-coils each consisting

of 300-400 aa. Coiled-coils of the arm domains have

a number of defects in which one or both antiparal-

lel chains lose their regular α-helical structure [19].

Two such breaks in the coiled-coil of the arm domain

have high degree of evolutionary conservation and,

most likely, play an important role in the activity of

SMCcomplexes. The first one, located around the mid-

dle of the arm domain a bit closer to the hinge domain

has been called ‘elbow’ (Fig. 2a). This region ensures

mechanical flexibility of the arm domains: simulta-

neous bending of the two ‘elbow’ regions could bring

the hinge domains into the close proximity of the head

domains (Fig. 3b). The second conserved break of the

coiled-coil is located close to the head domain at the

distance of approximately 50aa; it is commonly called

‘joint’(Fig.2a). ‘Joints’ in the SMC subunits are import-

ant hubs of interactions with the auxiliary subunits

COHESIN COMPLEX STRUCTURE 589

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

ofthe complex [25], they also participate in the move-

ments of the head domains relative each other [26].

Head domains of the SMC proteins comprise ABC-

type ATPases[16,27]. Similarly to other ABC-proteins,

cohesin is capable of hydrolyzing ATP only during

physical interaction between ATPase domains of two

SMC subunits (Fig. 3a) [28, 29]. Head domain of one

of the subunit of the dimer binds an ATP molecule,

and the head domain of the other is required for hy-

drolysis of the bound molecule. Each cycle of the co-

hesin-dependent ATP hydrolysis involves binding of

ATP by each of the head domains, and as a result two

ATP molecules are hydrolysed. Engagement of head

domains is terminated after nucleoside triphosphate

hydrolysis and is established again after binding of a

new pair of ATP molecules. Hence, in contrast to inter-

action of the hinge domains with each other, dimeriza-

tion of the head domains depends on the presence of

the substrate and has dynamic nature (Fig.3a).

From the structural point of view the ABC-ATPase

of the head domain is a globule with the core consisting

of an open β-cylinder with β-strands belonging to both

N- and C-terminal parts of the SMC protein[9, 16,28].

At one of the poles of the β-cylinder there are two con-

served catalytically significant sequences: Walker A mo-

tif (also known as a P-loop or phosphate-binding loop)

and Walker B motif. Both these motifs are typical not

only for all representatives of the ABC-ATPase super-

family, but also for the wider monophyletic protein

group called P-loop NTPases[27]. The Walker A motif,

located in the N-terminal part, is responsible for bind-

ing β- and γ-phosphate groups of ATP. The Walker B

motif is located in the C-terminal part and is responsi-

ble for coordination of Mg

in the active center. Unlike

in the case of other P-loop NTPases, effective hydroly-

sis of the phosphate bond by the ABC-ATPases requires

spatial interaction between nucleotide triphosphate

bound by the Walker A and B motifs and the serine

residue within the so-called SGG-motif. The SGG-mo-

tif is also located in the head domain, but at the rela-

tively large distance from the ATP-binding β-cylinder,

hence, such interaction is impossible within the head

of SMC monomer, it could be realized only between

the ATP-binding site of one head domain and SGG-mo-

tif of another one during their dimerization (Fig. 3a).

That is why ATPase activity of cohesin can be realized

only during physical engagement of the head domains

of two SMC subunits of the complex.

The structured N- and C-terminal domains of

RAD21 interact with two head domains of two SMC

proteins asymmetrically (Fig.2b)[30]. The N-terminal

domain of RAD21 binds to the site on the arm coiled-

coil coming out of the head domain of SMC3 Part of

the arm domain that binds N-terminal domain of klei-

sin is located between the head domain and ‘joint’

and is called ‘neck’. Fragment of the neck site of the

SMC3 coiled-coil domain interacts with two N-terminal

α-helices of RAD21 with formation of four-α-helix bun-

dle[30,31]. The C-terminal globular domain of RAD21

has a winged-helix fold and interacts with the top of

the SMC1 head domain[32].

INTERACTIONS

OF THE REGULATORY HAWK-PROTEINS

WITH THE KLEISIN

The core cohesin trimer interacts with three

auxiliary subunits: STAG1/2

 (Scc3), PDS5A/B (Pds5),

and NIPBL (Scc2). The auxiliary subunits participate

in the complex binding to DNA and regulation of its

activity [1-3]. The disordered central part of RAD21

represents the major binding site for the auxiliary sub-

units on the core trimer [33-35].

All three auxiliary subunits of cohesin belong to

the family of HAWK-proteins (HEAT proteins associat-

ed with kleisin) [15]. These α-helical hook-shaped pro-

teins (Fig. 2c) [33, 36] are composed of around 20tan-

dem HEAT-repeats[37]. Each repeat consists of 30-40aa

organized as a pair of interacting antiparallel amphip-

athic α-helices. α-Helices of each HEAT- repeat are ar-

ranged perpendicularly to the axis of the HAWK-sub-

unit hook. Inner concave surface of the HAWK hook

interacts with kleisin-subunit.

STAG1/2 binds to kleisin stably and, hence, is a

constitutive component of the complex [38]. NIPBL and

PDS5A/B interact with kleisin transiently; moreover,

these two subunits compete with each other for

the same binding site [35]. Hence, at each time point

the complex could contain one (STAG1/2) or two

(STAG1/2+ NIPBL or STAG1/2 + PDS5A/B) HAWK-sub-

units (Fig. 2b). Binding of NIPBL or PDS5A/B dramat-

ically changes activity of the cohesin complex [35].

NIPBL increases substantially ATPase activity of the

complex [39, 40]. This subunit is essential for the co-

hesin-dependent loop extrusion[40, 41]. PDS5A/B-sub-

unit supresses ATPase activity, and in its presence the

WAPL protein could be recruited to the cohesin ring,

which stimulates dissociation of the N-terminal part

of RAD21 from the SMC3 subunits[42, 43]. Acetylation

of the SMC3 subunit, as well as interaction of the com-

plex with other chromatin components such as insu-

lator protein CTCF, regulates binding of NIPBL and

PDS5A/B to the cohesin ring[25, 44-46].

Vertebrate genomes generally encode a pair of somatically expressed paralogs for both Scc3 and Pds5 HAWK-proteins:

STAG1/STAG2 and PDS5A/PDS5B. These paralogs in the majority of cases are structurally and functionally equivalent, hence,

here and further in the text the designations STAG1/2 and PDS5A/B are used.

GOLOV, GAVRILOV590

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

CONFORMATIONAL POLYMORPHISM

OF COHESIN COMPLEX

Activity of all proteins one way or another is asso-

ciated with ligand-dependent conformational changes.

SMC complexes demonstrate ability of radical confor-

mational rearrangements, which mediate their motor

activity.

Various microscopic techniques, X-ray crystallog-

raphy, as well as indirect approaches for elucidation of

structure such as Förster Resonance Energy Transfer

(FRET), chemical cross-linking, and analysis of stability

of DNA–protein complexes in various physicochemical

conditions facilitated discovery of a wide spectrum of

cohesin conformational states (Fig.3b)[3,4].

Entire variety of the known cohesin conforma-

tions could be described to a large degree by three

parameters: (1) engagement of head domains of the

two SMC subunits, (2) distance between the hinge and

head domains controlled by the degree of the elbow

bending, (3) distance between the arm coiled-coils of

two SMC subunit. For example, in both I-conformation

and folded conformation head domains are engaged,

and arm’s coiled coils interact along the entire length;

difference is in the fact that in the first conformation

arm domains are in the straightened state, while in

the second conformation – in the fully bended state.

O- and B-conformations differ from the previous pair

by the fact that interactions between the arm domains

are absent (Fig.3b).

Part of conformational rearrangements of cohes-

in is strictly coupled with ATP binding or hydrolysis,

while others occur spontaneously. It is likely though

that the direction of some spontaneous cohesin rear-

rangements could be controlled by the irreversible

processes of ATP binding and hydrolysis.

Dimerization of head domains and ATP hydro-

lysis. ATP binding induces tight interaction between

the two cohesin head domains, the so-called E-state

(engaged state) [9, 28, 47]. This short-lived structure

breaks down after ATP hydrolysis (Fig.3a). ATP hydro-

lysis is followed by the release of the products of hy-

drolysis and change in positioning of the head domain

relative to each. It was originally assumed that ATP

hydrolysis inevitably leads to a complete separation of

head domains in a structure known as an O-state (open

state). However, it has been later discovered that even

in the absence of ATP the structure called J-state (jux-

taposed state) can be formed in which two head do-

mains interact in such a way that the SGG-motifs of the

pair of SMC proteins are brought close to each other

[26, 48] (Fig. 3a). Considering that this interaction is

realized through the surfaces located at a relatively

large distance in the E-state, transitions from E-state

to J-state and back involve significant rotation of the

head domains relative to each other.

Additional conformational changes coupled

with binding and hydrolysis of ATP. ATP binding

and engagement of head domains in E-state is coupled

with several important conformational rearrangements:

local separation of arm domains, temporal dissocia-

tion of kleisin from SMC3, establishing of an addition-

al bridge between the two head domains formed by

NIPBL[25,44,49].

Coiled-coils of the arm domains that are adjacent

to the head domains are in fixed orientation relative

to each other in the J- and E-states. Despite the relative

flexibility of the arm domains, comparatively long re-

gions of the coiled coils adjacent to the head domains

are constitutively separated in the E-state [25, 44, 49].

At the same time, position of the head domains in the

J-state ensures close proximity of the coiled-coil re-

gions of SMC1 and SMC3[26, 48]. The available struc-

tural data indicate that the arm domains in the J-state

are in contact with each other over their entire length

(Fig.3a). Hence, while in the absence of ATP (in the so-

called apo-form) the arm domains could be either sep-

arated, or interact with each other to a certain degree,

binding of ATP results in separation of the coiled coils

at least in the regions directly adjacent to the head do-

mains.

During interaction of the head domains with each

other in the E- and J-state the united S-K-compartment

is divided into two compartments (SandK) confined by,

respectively, SMC dimer and kleisin subunit (Fig. 3a).

Depending on the nature of the interaction between

the head domains, the following subcompartments can

be distinguished: E-K, E-S, J-K, and J-S; due to the tight

contact between the coiled coils of the arm domain,

the latter does not have any opening.

Another change of cohesin conformation associat-

ed with ATP binding in the active center is short-term

opening of the so-called N-kleisin gates of the complex.

At the moment of formation of the E-state, the N-termi-

nal domain of RAD21 dissociates from its binding site

at the SMC3 subunit[47,49], which results in temporal

disruption of integrity of the E-K compartment.

ATP binding also causes changes in the pattern

of NIPBL interaction with other subunits. In addition

to its main binding site in the central part of kleisin,

NIPBL has additional ATP-regulated sites of interac-

tion with the SMC subunits. In the absence of ATP

NIPBL interacts with the dimer of hinge domains;

ATP binding results in dissociation of NIPBL from the

hinge domains and attachment to the head domain of

SMC3 [50]. This jump of NIPBL, likely, stimulates for-

mation of the E-state, in which NIPBL forms addition-

al contacts with the head domain of SMC1[25, 44,49].

Apair of conserved lysine residues in SMC3, K105/K106

(K112/K113– inSmc3 S. cerevisiae), plays an important

role in the interaction of NIPBL with the SMC3 head

domain, acetylation of these residues by acetyl trans-

COHESIN COMPLEX STRUCTURE 591

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

ferase ESCO2(Eco1) is a key factor of stabilization of

the cohesive binding. Following ATP hydrolysis NIPBL

loses its interaction with the head domains and re-

stores interaction with the hinge domain site [50].

Spontaneous changes of cohesin conformation.

Microscopic observations, as well as the FRET data

indicate that the arm domains of cohesin and oth-

er SMC complexes are rather labile structures most-

ly due to the possibility of bending at elbows [17, 19,

49-51]. Amplitude of this bending reaches up to 180°

which makes possible direct physical interaction of the

head and hinge domains in the completely bent state

(Fig.3b).

Discovery of the J-state and recognition of the fact

that transitions between the E- and J-states should be

accompanied by the significant changes in the mutual

position of the arm domains led to the suggestion of a

hypothesis according to which bending and straighten-

ing of the elbow domains should be strictly controlled

by the ATP hydrolysis cycle [19]. However, experi-

mental data obtained later indicate that bending and

straightening of the elbows occur spontaneously [50].

The most important consequence of bending and

straightening of the elbows is the engagement of the

hinge domains with the head domains in the complete-

ly bent state. The possibility of temporal interaction be-

tween the head and hinge domains is a crucial premise

of many mechanistic models of SMC-dependent loop

extrusion[19, 50, 52, 53]. Despite the fact that bending

of the arm domains occurs spontaneously, there are

indication that there could be a ratchet mechanism

coupling bending and ATP hydrolysis. In particular, it

has been suggested that the complete bending of the

elbow domain could be required for the formation of

the E-state[25,44,49,50].

Another known spontaneous change of the cohes-

in architecture is reversible interaction of the coiled-

coils adjacent to the hinge domains. Such interaction

spreads from the hinge domain towards the elbows

and, likely, is associated with winding of the two

coiled-coil domains of the arm around each other [50].

It is hypothesized that this winding can be a necessary

prerequisite for the bending of the elbows.

INTERACTIONS BETWEEN DNA AND COHESIN

Binding to DNA is a key aspect of activity of SMC

complexes. There are two characteristic features of

interactions between SMC complexes and DNA: (1) as-

sociation between changes in affinity of different sites

to DNA and changes in complex conformation and

(2) ability of the complexes to entrap DNA topologically.

Unlike the classic electrostatic interactions typical

for the majority of DNA-binding proteins, topological

binding represents a qualitatively different mode of

non-covalent binding for which there is no need in

direct physical contact between the DNA and the pro-

tein [1, 44]. Topologically entrapped DNA is threaded

through the closed S-K-ring of the complex like a string

threaded through a bead (Fig. 4a). In this scenario

complex can be stably associated with DNA even in the

absence of electrostatic interactions.

Cohesin can also interact with DNA electrostat-

ically: several positively charged sites on the com-

plex surface mediate such interactions (Fig.5)[3, 50].

An important feature of the electrostatic interactions

of cohesin with DNA is cooperativity: small individu-

al sites dispersed over the different subunits are not

generally capable to form stable contacts with DNA on

their own, sufficient affinity is achieved only by the

composite surfaces formed by several DNA-binding

sites from different subunits coming together. Such

composite surfaces are assembled in some cohesin con-

formations and disassembled in others. Thedynamical-

ly assembled DNA-binding sites are crucial elements of

the machinery that couples the processes of ATP hydro-

lysis, structural changes of the complex, and its move-

ment along the DNA thread during loop extrusion.

Topological entrapment of DNA within cohesin.

Cohesin, same as almost all other SMC complexes, is

capable of binding DNA topologically

. Three observa-

tions underlie the ring hypothesis according to which

cohesin could entrap DNA topologically inside the

three- part S-K-ring [22, 54]: (1) complex structure sug-

gesting existence of the extended topologically closed

compartment [17, 55]; (2) capability of the complex to

establish extremely stable interactions with chromatin

during the G2-phase[56, 57], and (3)  immediate desta-

bilization of these interactions upon proteolysis of the

kleisin subunit in anaphase [21, 58]. After more than

a decade of accumulation of indirect clues, this hy-

pothesis was finally confirmed with the experiments

including covalent cross-linking of the cohesin rings

invivo[7,30,48].

A number of experimental approaches have been

suggested for confirming topological nature of protein

and protein complexes binding to DNA. Topological

nature of interaction may be indicated by the follow-

ing experimental observations: (i) stability of binding

in the high ionic strength buffers[39,56] (Fig.4b(1));

(ii) sensitivity of the interaction to proteolytic cleavage

In certain conformations cohesin can form, in addition to S-K ring, other non-covalently closed compartments in which DNA

strand can be entrapped. Such structures include NIPBL-SMC3-subcompartment, E-S-, E-K-, and J-K-subcompartments. Integ-

rity of such subcompartments, unlike integrity of the S-K-ring, is usually very quickly disrupted. In the text term ‘topological

binding’ refers to the DNA entrapment within the stable S-K-ring.

GOLOV, GAVRILOV592

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

Fig. 4. Topological interaction of the cohesin ring with DNA. a)Topological(1) and pseudo-topological(2) interaction of cohesin

with DNA. In the pictograms direction of DNA-thread relative to the figure plane is shown with●and+ symbols. b)Methods used

to establish topological nature of interaction between the protein complex and DNA: analysis of stability of the binding in the

high ionic strength buffers(1); analysis of sensitivity of interaction to proteolytic cleavage of one of the subunits(2); analysis

of sensitivity of interaction to the break in the DNA molecule(3); analysis of the interaction stability under denaturing condi-

tions after covalent cross-linking of the protein ring with the help of cysteine specific cross-linking agents (BMOE,bBBr)(4).

c)Two pathways for removal of topologically bound cohesin rings from DNA realized in eukaryotic cells: WAPL-dependent(1)

and proteolytic(2).

of one of the subunits of the complex [22, 39]

(Fig. 4b (2)); (iii)  sensitivity to double-strand DNA

break (if circular DNA molecule participates in the

interaction)[39, 59] (Fig.4b(3)); (iv) resistance of the

interactions to denaturing conditions after covalent

cross-linking of the protein ring [30,60] (Fig.4b(4)).

Conclusions made based on the results of experiments

using first three of the mentioned approaches are

mostly of preliminary nature, while the experiments

involving cross-linking of the protein ring allow con-

cluding with confidence on the mode of the complex

binding to DNA.

COHESIN COMPLEX STRUCTURE 593

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

Fig. 5. Electrostatic interactions between cohesin and DNA. a)Two hypothetical pathways of the formation of ‘DNA gripping’

structure suggested based on observations of S.cerevisiae(1) and Schizosaccharomyces pombe(2) cohesins. In the pictograms

direction of DNA-thread relative to the figure plane is shown with ● and + symbols, gray fill indicates topological binding.

In the pairs of pictograms, the left one reflects the formation of E-S- and E-K-subcompartments and position of DNA threads rel-

ative to them; the right one– position of DNA threads relative to the S-K-ring. Dashed parts of the RAD21-subunit correspond to

the regions in which path of the protein chain is shown arbitrary for clarity (in reality HAWK-subunits remain bound to RAD21

at all presented stages). b)Electrostatic interactions of the cohesin hinge domains with DNA. Three (not mutually exclusive)

scenarios are shown: NIPBL-subunit-mediated interaction(1), direct contact of the inner surface of the hinge domain pore with

DNA(2), and direct contact of DNA with the southern pole of the hinge domains dimer(3). c)Interaction of HAWK-B subunits

of cohesin (STAG1/2) and condensin (CAP-G) with DNA. Peptide loop, called ‘safety belt’, formed by the kleisin subunit of conden-

sin (CAP-H), additionally stabilizes binding of the HAWK-B subunit(CAP-G) to DNA.

The most reliable proofs of the ability of SMC

complexes to interact with DNA topologically were

obtained during investigation of the covalently cross-

linked S-K-rings [7, 60, 61]. Resolution of the atomic

structure of three sites of interactions between the

pairs of the subunits forming S-K-ring (dimer of the

SMC1 and SMC3 hinge domains, N-terminal RAD21

domain interacting with the neck region of SMC3, and

C-terminal domain of RAD21 interacting with the SMC1

head domain) allowed to genetically engineer cohesin

complexes so that the S-K-compartments can be cova-

lently crosslinked with the bifunctional thiol-specific

reagents: bis-maleimidoethane (BMOE) or dibromobi-

mane (bBBr). Capability of complex to be cross-linked

is achieved by introduction of three pairs of cysteine

residues at the sites of interaction of three core sub-

units in such a way that two cysteines in each pair

are located opposite to each other on different sub-

units of the complex at a distance no larger than 1nm

(Fig. 4b (4)). Long linear molecules of genomic DNA

and circular plasmids topologically interacting with

the modified cohesin complexes cannot be separated

after treatment with the cross-linking agents from the

protein trimer even under the harsh denaturing condi-

tions, which allows easy detection of such DNA mole-

cules by the decrease in their electrophoretic mobility.

Existence of cohesin complexes topologically interact-

ing with DNA not only in artificial in vitro systems,

GOLOV, GAVRILOV594

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

but also in living cells has been finally proved with the

help of such approach [7, 30,48]. Besides detection of

the mere fact of topological interaction thiol-specific

cross-linking of other pairs of artificially introduced

cysteines potentially allows establishing the identity of

specific subcompartment of the S-K-ring (E-S, E-K, J-S,

or J-K) involved in topological interaction between co-

hesin and DNA (Fig.3a)[44,48,62].

In order to understand physiological activity of

cohesin it is important to investigate mechanisms of

formation and termination of topological interactions

of the S-K-ring with DNA. Theoretically such interac-

tion could be established and terminated during tem-

poral disengagement of one of the three non-covalent

interfaces forming the S-K-ring: interface between

SMC3 and SMC1 hinge domains (hinge gates), interface

between kleisin and SMC3 subunit (N-kleisin gates), or

interface between kleisin and SMC1 (C-kleisin gates)

(Fig. 4a (1)). In addition, topological interaction of co-

hesin with DNA could be disrupted due to proteolysis

of one of the core subunits of the complex.

Mechanisms of termination of the topological

binding of cohesin are relatively well understood.

It has been established that non-proteolytic removal

of the complex from DNA (Fig.4c(1)) is catalyzed by

the conserved nuclear protein WAPL (Wpl1) (winged-

apart like) and is caused by the opening of N-kleisin

gates [42, 63, 64]. The structured C-terminal domain

of WAPL consists of eight HEAT-repeats and it can

interact with the engaged SMC head domains in the

E-state. The non-structured N-terminal domain con-

tains conserved YSR- and FGF-motifs participating in

the interaction with the HAWK-subunits: PDS5A/B and

STAG1/2[43]. WAPL is not a constitutive component of

the cohesin complex, but can transiently interact with

the PDS5A/B-containing rings [42, 65]. It is assumed that

the WAPL binding and DNA release occur when the

complex is in the E-state. Formation of the E-state al-

ways results in the short-term opening of the N-kleisin

gates, however, WAPL stabilizes the complex in such

open state, which increases the probability of DNA

leaving the ring [43,66]. According to this mechanism

in order for the DNA to escape the ring it should be

first transferred into the E-K-subcompartment. After

the DNA release from the complex, dissociation of WAPL

from the complex and closing of exit gates occur.

Topological interaction of cohesin with DNA is dis-

rupted also due to proteolysis of the kleisin subunit in

anaphase [21, 58,67]. After the passage of the anaphase

checkpoint the serine protease separase cleaves the

central non-structured part of the RAD21-subunit and,

thus, releases the sister chromatids from the topologi-

cal entrapment within the cohesin ring (Fig.4c(2)).

Much less is known about how topological binding

of cohesin to DNA is established. Ability of cohesin to

hydrolyse ATP as well as presence of NIPBL and STAG1/2

are required for topological loading of the complex

in cells and in in vitro systems [7, 44]. At the same

time, other data indicate that at least in vitro the

PDS5A/B–WAPL complex could catalyze reaction of co-

hesin loading onto DNA in the absence of the NIPBL-

subunit[39,68].

Not only the details of the loading mechanism are

still unknown, but also what particular gates of the

SMC complex let DNA inside the ring. On the one hand,

it has been assumed that formation of the E-state is

accompanied by the transient opening of the N-klei-

sin gates through which the DNA thread could enter

the S-K-compartment [49]. On the other hand, it was

shown that the complexes with covalently closed N-

or C-kleisin gates despite all that can be topologically

loaded onto DNA [7]. Moreover, it has been shown that

the positively charged amino acids located inside the

pore between the hinge domains are required for es-

tablishing topological interactions between the cohes-

in and DNA, which indicates possible role of the hinge

gates in this process[7,8].

From the early days of the hypothesis suggest-

ing topological mode of cohesin–DNA interactions it

was assumed that the S-K-ring can accommodate two

DNA strands simultaneously ensuring cohesion of sis-

ter chromatids in the G2-phase of cell cycle [22, 54].

Later this mechanism of cohesion was confirmed ex-

perimentally [7, 48]. Some of the currently discussed

models of cohesin-dependent loop extrusion suggest

that the base of growing loop consists of two DNA

strands entrapped within a single S-K-ring[26, 53,62].

Such loop can be formed while the S-K-compartment

remains constantly closed, it does not contain true

topological link between the SMC complex and DNA,

and therefore it represents an example of the so-called

pseudo-topological interactions (Fig.4a(2)). Unlike the

topological entrapment of DNA within cohesin, pseudo-

topological interaction is still considered to be a specu-

lative model rather than the observed phenomenon.

The topologically engaged cohesin complexes are

capable of passive one-dimensional diffusion along

the DNA thread [23, 24]. In vitro studies showed that

the size of cohesin pore allows this complex to by-

pass small DNA-bound proteins during such diffu-

sion, however, nucleosome particles with diameter of

10 nm represent a significant obstacle, while particles

with diameter 20 nm are impassable barriers for such

motion [23]. It has been suggested that the pseudo-to-

pologically loaded cohesin rings are also capable of

diffusion along the DNA threaded through the com-

plex pore[40,53].

DNA gripping: transient ATP-dependent bind-

ing of head domains and NIPBL-subunit to DNA.

The most important finding in structural biology of

cohesin was the discovery of coupling between the for-

mation of E-state in the presence of ATP and cohesin

COHESIN COMPLEX STRUCTURE 595

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

binding to DNA. It was observed that the cohesin com-

plexes with EQ-mutations in the Walker B motif (sup-

pressing ATPase activity) in the presence of ATP, as

well as wild type cohesin complex in the presence of

nonhydrolyzable ATP analogues form stable complex-

es with DNA called gripping state/DNA clamping [25,

44, 49]. Presence of NIPBL, which is not a constant

component of cohesin, is also required for formation

of such complex. During the gripping double helix

interacts with the dimerized head domains (Fig. 5a),

moreover, DNA is found pinned to the SMC3 head do-

main with the help of NIPBL, which in the presence of

ATP binds to the joint region of the SMC3 arm domain

and head domains of both SMC subunit.

In the gripping configuration the DNA sugar-phos-

phate backbone forms a series of electrostatic interac-

tions with different cohesin sites: upper surface of the

dimerized SMC head domains, DNA-binding surface

of NIPBL and N-terminal domain of RAD21. All DNA

binding sites participating in DNA gripping come close

to each other and form a small channel complemen-

tary to the bound DNA double helix only for a short

time, while the complex is in the gripping state. After

ATP hydrolysis, the E-state is disassembled, and at the

same time the composite DNA-binding surface is taken

apart. Hence, in the wild type complexes, gripping is a

short-lived state, formation and disruption of which is

coupled with ATP binding and hydrolysis.

In the presence of ATP, DNA, NIPBL, and wild type

cohesin the head SMC domains and NIPBL-subunits

periodically interact with each other; replacement of

ATP with nonhydrolyzable analogue or introduction

of EQ-mutations renders this interaction constitutive

[49,50]. Remarkably, on binding of ATP, the DNA first

interacts with the SMC3 head domain and NIPBL-sub-

unit in the so-called pre-engaged clamp, and only af-

ter this engagement of the head domains occur with

formation of the full-fledged gripping configuration

(Fig.5a(1))[50].

In the absence of DNA, the head SMC domains

are capable of dynamic interactions with fast forma-

tion and disruption of the E-state [49, 50]; addition

of DNA to the system does not increase frequency of

such interactions [49, 50], however, it does stimulate

significantly ATPase activity of cohesin [39, 40]. Most

likely, DNA increases productivity of the interactions

between the head domains due to allosteric effect on

the active center which makes ATP hydrolysis a neces-

sary condition for disengagement of the head domains

[50]. This explains why catalytically inactive (contain-

ing EQ-mutations) SMC complexes in the presence of

DNA and ATP remain frozen in the gripping configu-

ration [44,49].

Two alternative pathways were suggested for the

formation of gripping state. The data obtained with

the help of thiol-specific cross-linking of S. cerevisiae

cohesin show that during formation of gripping state

DNA thread is trapped simultaneously in both E-K-

and E-S-subcompartments. However, DNA does not

enter the S-K-compartment; hence, it is assumed that

in this case the S-K-ring harbors a small DNA loop in

a pseudo-topological manner (Fig. 5a (1)) [44]. It has

been suggested that formation of such complex oc-

curs due to the approaching of DNA to cohesin from

the side of disengaged head domains, and passing

through them prior to formation of the E-state makes

DNA trapped simultaneously in the E-K- and E-S-sub-

compartments (Fig. 5a (1)). As a result, formation of

the gripping state according to this mechanism is not

accompanied by DNA topological entrapment within

the S-K-ring. The data on the kinetics of DNA entrance

into the E-K- and E-S-subcompartments are in agree-

ment with the proposed mechanism: short incubation

of DNA with cohesin is sufficient for realization of

these reactions, moreover both reactions follow the

same kinetics.

The results obtained during the study of DNA-

gripping state formation by the Schizosaccharomyces

pombe cohesin contradict the scenario described

above [49]. These data demonstrate that first small

DNA loop is threaded through the S-K-ring, this is fol-

lowed by the formation of the gripped state with the

lower part of the loop located closer to the head do-

mains being clamped (Fig. 5a (2)). It is suggested that

while the head domains establish E-state interaction

the N-kleisin gates open temporarily and could occa-

sionally let DNA within the S-K-ring. Thus, according

to this model the DNA gripping state can be one of

the intermediates of the process of topological load-

ing. However, in the framework of this model the

DNA thread usually does not enter the N-kleisin gates

during formation of the gripping state, and thus the

majority of the ATP hydrolysis cycles do not result in

the topological loading of the complex.

Deciphering process of DNA gripping state forma-

tion is crucial for understanding of the mechanisms of

cohesin-dependent loop extrusion and of the relation-

ships between loop extrusion and topological entrap-

ment of DNA within the S-K-rings [4,5,53].

Electrostatic interactions of hinge domains

with DNA. The dimerized hinge domains of the cohes-

in SMC subunits interact with double-stranded DNA

electrostatically [50, 69, 70]. In the absence of ATP (in

the apo-form of the complex), hinge domains interact

with NIPBL[50], which results in formation of a stron-

ger binding site combining DNA-binding surfaces of the

hinge domains and of the HAWK-subunit (Fig. 5b (1)).

Binding of ATP by the head domains results in disrup-

tion of this composite site and decrease in affinity of

the hinge domains to DNA.

There is no reliable information on particular

amino acids in the hinge domains that participate in

GOLOV, GAVRILOV596

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

the DNA binding (Fig. 5b (2 and 3)). Despite the fact

that the positively charged amino acids lining the

small pore of the dimer of hinge domains are re-

quired for the DNA entrapment within the S-K-com-

partment [7, 8], their physical interaction with DNA

has not been demonstrated yet. The pore is too small to

trap the double-stranded DNA within it; nevertheless,

some authors suggest that the loss of contacts between

the hinge domains at one pole of the dimer could ex-

pose inner surface of the pore for DNA binding [7,71]

(Fig.5b(2)). Interestingly enough, one of the published

structures of the cohesin complex in the gripping con-

figuration contains hinge domains with disengaged

northern dimerization surface [25].

According to other data, the sites on the hinge

domains located at the southern pole of the dimer

looking inside the S-K-ring as well as adjacent re-

gions of the coiled coils are responsible for DNA bind-

ing[50,70](Fig.5b(3)).

Participation of STAG1/2 in cohesin–DNA in-

teraction. Unlike NIPBL, which, most likely, contacts

DNA exclusively as a constituent of the composite

DNA-binding sites in cooperation with the head or

hinge domains, STAG1/2 is more or less autonomous

DNA-binding module (which, however, does not rule

out its participation in cooperative binding; in partic-

ular, there are indications that STAG1/2 interacts with

DNA in the gripping state in close proximity of the

site of DNA interaction with the NIPBL-subunit) [25].

Amino acid residues responsible for interaction of

STAG1/2 with DNA are localized in the regions ho-

mologous to the corresponding DNA-binding patches

on the NIPBL surface (Fig. 5c). One of the main such

regions is the positively charged groove at the side

surface of the long N-terminal subdomain of the

hook-shaped molecule [50, 72]. Interestingly enough,

electrostatic interaction of the CAP-G/CAP-G2-subunit

of condensin – STAG1/2 homolog – with DNA is addi-

tionally stabilized due to the fact that the kleisin sub-

unit CAP-H/CAP-H2 forms a non-covalent peptide loop

around the DNA thread interacting with CAP-G/CAP-G2

[62, 73, 74] (Fig. 5c). This loop with length of around

100 aa is called ‘safety belt’ and is formed as a result

of interaction between the two short hydrophobic frag-

ments in the middle of the non-globular part of the

kleisin. The kleisin cohesin subunit, RAD21, most likely,

does not have a structure similar to the condensin’s

‘safety belt’.

CONCLUSIONS

Some structural features of the SMC complexes

such as ring architecture and their ability to topolog-

ical entrap DNA were described long before the dis-

covery of the loop extrusion process. Exploration of

cohesin ability to topologically entrap DNA was cou-

pled with elucidation of molecular mechanisms of es-

tablishing, maintenance, and termination of cohesion.

However, cohesion, which is not directly associated

with extrusion activity, is an exception rather than the

rule: vast majority of physiological activities of cohes-

in and other SMC complexes depend on their ability to

actively create DNA loops. The fact that these complex-

es were found to be motor proteins spurred further in-

terest in their structure.

Aspiration to decipher molecular mechanism of

extrusion motivated researchers to investigate struc-

ture of cohesin in recent years. Breakthroughs in this

area are at large associated with the development of

experimental technologies that enable exploration of

dynamic and, to a certain degree, polymorphic struc-

ture of massive multisubunit SMC complexes. Technol-

ogies that substantially contributed to the current un-

derstanding of the cohesin structure include cryogenic

electron microscopy [25, 44,49], atomic force micros-

copy [50], FRET [49, 50], and protocols for real-time

imaging of loop extrusion in reconstituted invitro sys-

tems [8,40]. These methods unraveled many disparate

molecular details of the extrusion process. One of the

most important findings in this respect was the discov-

ery of the periodical formation of cohesin DNA-grip-

ping state during the loop extrusion process, as well as

establishing of the central role of NIPBL in the assem-

bly of this structure.

Unfortunately, disparate structural, biochemical,

and genetic observations still do not provide a com-

prehensive model describing the process of cohes-

in-dependent loop extrusion (hypothetical mechanistic

models of this process based on the available data on

the structure and activity of SMC complexes are pre-

sented in the second part of this review [10]). Never-

theless, there is every reason to believe that accumu-

lation of empirical, primarily structural data could

facilitate elucidation of the molecular mechanism of

SMC-dependent loop extrusion in the nearest future.

Acknowledgments. The authors are grateful to

S. V. Razin for fruitful and detailed discussion of nu-

merous topics addressed in this review. The authors

also express their gratitude to A. V. Golova for the help

in preparation of illustrations.

Contributions. A.K.G. summarizing of the avail-

able data, writing first draft of the paper; A.A.G. for-

mulation of the problem, supervision of the work, and

editing of the paper.

Funding. This work was financially supported by

the Russian Science Foundation (project no.21-64-00001).

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.

COHESIN COMPLEX STRUCTURE 597

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

REFERENCES

1. Yatskevich, S., Rhodes, J., and Nasmyth, K. (2019)

Organization of chromosomal DNA by SMC com-

plexes, Annu. Rev. Genet., 53, 445-482, doi: 10.1146/

annurev-genet-112618-043633.

2. Gligoris, T., and Löwe, J. (2016) Structural insights

into ring formation of cohesin and related Smc com-

plexes, Trends Cell Biol., 26, 680-693, doi: 10.1016/

j.tcb.2016.04.002.

3. Lee,H., Noh, H., and Ryu, J.-K. (2021) Structure-func-

tion relationships of SMC protein complexes for

DNA loop extrusion, Biodesign, 9, 1-13, doi:10.34184/

kssb.2021.9.1.1.

4. Davidson, I.F., and Peters, J.-M. (2021) Genome fold-

ing through loop extrusion by SMC complexes, Nat.

Rev. Mol. Cell Biol., 22, 445-464, doi: 10.1038/s41580-

021-00349-7.

5. Oldenkamp, R., and Rowland, B. D. (2022) A walk

through the SMC cycle: from catching DNAs to shap-

ing the genome, Mol. Cell, 82, 1616-1630, doi:10.1016/

j.molcel.2022.04.006.

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

Popov, A., Salnikov, P., Battulin, N., and Fishman, V.

(2023) Function and evolution of the loop extru-

sion machinery in animals, Int. J. Mol. Sci., 24, 5017,

doi:10.3390/ijms24055017.

7. Srinivasan, M., Scheinost, J. C., Petela, N. J., Gligor-

is, T.G., Wissler,M., Ogushi,S., Collier, J. E., Voulgar-

is,M., Kurze,A., Chan, K.-L., Hu,B., Costanzo,V., and

Nasmyth, K.A. (2018) Thecohesin ring uses its hinge

to organize DNA using non-topological as well as

topological mechanisms, Cell, 173, 1508-1519.e18,

doi:10.1016/j.cell.2018.04.015.

8. Nagasaka, K., Davidson, I. F., Stocsits, R. R.,

Tang, W., Wutz, G., Batty, P., Panarotto, M., Litos, G.,

Schleiffer, A., Gerlich, D. W., and Peters, J.-M. (2023)

Cohesin mediates DNA loop extrusion and sis-

ter chromatid cohesion by distinct mechanisms,

Mol. Cell, 83, 3049-3063.e6, doi: 10.1016/j.molcel.

2023.07.024.

9. Arumugam,P., Nishino,T., Haering, C.H., Gruber,S.,

and Nasmyth, K. (2006) Cohesin’s ATPase activity

is stimulated by the C-terminal winged-helix do-

main of Its kleisin subunit, Curr. Biol., 16, 1998-2008,

doi:10.1016/j.cub.2006.09.002.

10. Golov, A.K., and Gavrilov, A.A. (2024) Cohesin-depen-

dent loop extrusion: molecular mechanics and role in

cell physiology, Biochemistry (Moscow), 89, 601-625,

doi:10.1134/S0006297924040023.

11. Nasmyth,K., and Haering, C.H. (2005) Thestructure

and function of SMC and kleisin complexes, Annu.

Rev. Biochem., 74, 595-648, doi: 10.1146/annurev.

biochem.74.082803.133219.

12. Cobbe,N., and Heck, M.M.S. (2004) Theevolution of

SMC proteins: phylogenetic analysis and structural

implications, Mol. Biol. Evol., 21, 332-347, doi:10.1093/

molbev/msh023.

13. Schleiffer,A., Kaitna,S., Maurer-Stroh,S., Glotzer,M.,

Nasmyth, K., and Eisenhaber, F. (2003) Kleisins: a

superfamily of bacterial and eukaryotic SMC pro-

tein partners, Mol. Cell, 11, 571-575, doi: 10.1016/

s1097-2765(03)00108-4.

14. Harvey, S.H., Krien, M.J.E., and O’Connell, M.J. (2002)

Structural maintenance of chromosomes (SMC) pro-

teins, a family of conserved ATPases, Genome Biol.

, 3,

REVIEWS3003, doi:10.1186/gb-2002-3-2-reviews3003.

15. Wells, J. N., Gligoris, T. G., Nasmyth, K. A., and

Marsh, J. A. (2017) Evolution of condensin and co-

hesin complexes driven by replacement of Kite by

Hawk proteins, Curr. Biol., 27, R17-R18, doi: 10.1016/

j.cub.2016.11.050.

16. Löwe,J., Cordell, S.C., and van den Ent,F. (2001) Crys-

tal structure of the SMC head domain: an ABC ATPase

with 900 residues antiparallel coiled-coil inserted,

J.Mol. Biol., 306, 25-35, doi:10.1006/jmbi.2000.4379.

17. Haering, C. H., Löwe, J., Hochwagen, A., and

Nasmyth, K. (2002) Molecular architecture of SMC

proteins and the yeast cohesin complex, Mol. Cell, 9,

773-788, doi:10.1016/S1097-2765(02)00515-4.

18. Kurze,A., Michie, K.A., Dixon, S.E., Mishra,A., Itoh,T.,

Khalid, S., Strmecki, L., Shirahige, K., Haering, C. H.,

Löwe,J., and Nasmyth,K. (2011) Apositively charged

channel within the Smc1/Smc3 hinge required for

sister chromatid cohesion, EMBO J., 30, 364-378,

doi:10.1038/emboj.2010.315.

19. Bürmann, F., Lee, B.-G., Than, T., Sinn, L., O’Reilly,

F.J., Yatskevich,S., Rappsilber,J., Hu,B., Nasmyth,K.,

and Löwe, J. (2019) A folded conformation of Muk-

BEF and cohesin, Nat. Struct. Mol. Biol., 26, 227-236,

doi:10.1038/s41594-019-0196-z.

20. Melby, T.E., Ciampaglio, C.N., Briscoe,G., and Erick-

son, H. P. (1998) The symmetrical structure of struc-

tural maintenance of chromosomes (SMC) and MukB

proteins: long, antiparallel coiled coils, folded at a

flexible hinge, J.Cell Biol., 142, 1595-1604, doi:10.1083/

jcb.142.6.1595.

21. Uhlmann,F., Wernic,D., Poupart, M.A., Koonin, E.V.,

and Nasmyth,K. (2000) Cleavage of cohesin by the CD

clan protease separin triggers anaphase in yeast, Cell,

103, 375-386, doi:10.1016/s0092-8674(00)00130-6.

22. Gruber, S., Haering, C. H., and Nasmyth, K. (2003)

Chromosomal cohesin forms a ring, Cell, 112, 765-777,

doi:10.1016/S0092-8674(03)00162-4.

23. Stigler, J., Çamdere, G. Ö., Koshland, D. E., and

Greene, E. C. (2016) Single-molecule imaging reveals

a collapsed conformational state for DNA-bound co-

hesin, Cell Rep., 15, 988-998, doi: 10.1016/j.celrep.

2016.04.003.

24. Davidson, I.F., Goetz,D., Zaczek, M.P., Molodtsov, M.I.,

Huis In ’t Veld, P.J., Weissmann,F., Litos,G., Cisneros,

D.A., Ocampo-Hafalla,M., Ladurner, R., Uhlmann,F.,

GOLOV, GAVRILOV598

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

Vaziri, A., and Peters, J.-M. (2016) Rapid movement

and transcriptional re-localization of human cohes-

in on DNA, EMBO J., 35, 2671-2685, doi: 10.15252/

embj.201695402.

25. Shi, Z., Gao, H., Bai, X.-C., and Yu, H. (2020) Cryo-EM

structure of the human cohesin-NIPBL-DNA complex,

Science, 368, 1454-1459, doi:10.1126/science.abb0981.

26. Diebold-Durand, M.-L., Lee, H., Ruiz Avila, L. B.,

Noh, H., Shin, H.-C., Im, H., Bock, F. P., Bürmann, F.,

Durand, A., Basfeld, A., Ham, S., Basquin, J., Oh, B.-

H., and Gruber,S. (2017) Structure of full-length SMC

and rearrangements required for chromosome or-

ganization, Mol. Cell, 67, 334-347.e5, doi: 10.1016/

j.molcel.2017.06.010.

27. Krishnan, A., Burroughs, A. M., Iyer, L. M., and Ara-

vind, L. (2020) Comprehensive classification of ABC

ATPases and their functional radiation in nucleo-

protein dynamics and biological conflict systems,

Nucleic Acids Res., 48, 10045-10075, doi: 10.1093/

nar/gkaa726.

28. Lammens, A., Schele, A., and Hopfner, K.-P. (2004)

Structural biochemistry of ATP-driven dimerization

and DNA-stimulated activation of SMC ATPases, Curr.

Biol., 14, 1778-1782, doi:10.1016/j.cub.2004.09.044.

29. Arumugam, P., Gruber, S., Tanaka,K., Haering, C.H.,

Mechtler, K., and Nasmyth, K. (2003) ATP hydroly-

sis is required for cohesin’s association with chro-

mosomes, Curr. Biol., 13, 1941-1953, doi: 10.1016/

j.cub.2003.10.036.

30. Gligoris, T. G., Scheinost, J. C., Bürmann, F., Pete-

la,N., Chan, K.-L., Uluocak,P., Beckouët,F., Gruber,S.,

Nasmyth,K., and Löwe, J. (2014) Closing the cohesin

ring: structure and function of its Smc3-kleisin in-

terface, Science, 346, 963-967, doi: 10.1126/science.

1256917.

31. Bürmann, F., Shin, H.-C., Basquin, J., Soh, Y.-M.,

Giménez-Oya,V., Kim, Y.-G., Oh, B.-H., and Gruber, S.

(2013) Anasymmetric SMC-kleisin bridge in prokary-

otic condensin, Nat. Struct. Mol. Biol., 20, 371-379,

doi:10.1038/nsmb.2488.

32. Haering, C. H., Schoffnegger, D., Nishino, T., Helm-

hart, W., Nasmyth, K., and Löwe, J. (2004) Structure

and stability of cohesin’s Smc1-kleisin interaction, Mol.

Cell, 15, 951-964, doi:10.1016/j.molcel.2004.08.030.

33. Roig, M. B., Löwe, J., Chan, K.-L., Beckouët, F., Met-

son,J., and Nasmyth,K. (2014) Structure and function

of cohesin’s Scc3/SA regulatory subunit, FEBS Lett.,

588, 3692-3702, doi:10.1016/j.febslet.2014.08.015.

34. Lee, B.-G., Roig, M.B., Jansma,M., Petela,N., Metson,J.,

Nasmyth,K., and Löwe, J. (2016) Crystal structure of

the cohesin gatekeeper Pds5 and in complex with

kleisin Scc1, Cell Rep., 14, 2108-2115, doi: 10.1016/

j.celrep.2016.02.020.

35. Petela, N.J., Gligoris, T.G., Metson,J., Lee, B.-G., Voul-

garis, M., Hu, B., Kikuchi, S., Chapard, C., Chen, W.,

Rajendra, E., Srinivisan, M., Yu, H., Löwe, J., and

Nasmyth, K.A. (2018) Scc2 is a potent activator of Co-

hesin’s ATPase that promotes loading by binding Scc1

withoutPds5, Mol. Cell, 70, 1134-1148.e7, doi:10.1016/

j.molcel.2018.05.022.

36. Ouyang, Z., Zheng, G., Tomchick, D. R., Luo, X., and

Yu, H. (2016) Structural basis and IP6 requirement

for Pds5-dependent cohesin dynamics, Mol. Cell, 62,

248-259, doi:10.1016/j.molcel.2016.02.033.

37. Neuwald, A. F., and Hirano, T. (2000) HEAT repeats

associated with condensins, cohesins, and other com-

plexes involved in chromosome-related functions, Ge-

nome Res., 10, 1445-1452, doi:10.1101/gr.147400.

38. Losada, A., Yokochi,T., Kobayashi, R., and Hirano, T.

(2000) Identification and characterization of SA/

Scc3p subunits in the Xenopus and human cohes-

in complexes, J. Cell Biol., 150, 405-416, doi: 10.1083/

jcb.150.3.405.

39. Murayama,Y., and Uhlmann,F. (2014) Biochemical re-

constitution of topological DNA binding by the cohesin

ring, Nature, 505, 367-371, doi:10.1038/nature12867.

40. Davidson, I.F., Bauer,B., Goetz,D., Tang,W., Wutz,G.,

and Peters, J.-M. (2019) DNA loop extrusion by human

cohesin, Science, 366, 1338-1345, doi:10.1126/science.

aaz3418.

41. Kim,Y., Shi,Z., Zhang,H., Finkelstein, I.J., and Yu,H.

(2019) Human cohesin compacts DNA by loop extru-

sion, Science, 366, 1345-1349, doi: 10.1126/science.

aaz4475.

42. Kueng, S., Hegemann, B., Peters, B. H., Lipp, J. J.,

Schleiffer, A., Mechtler, K., and Peters, J.-M. (2006)

Wapl controls the dynamic association of cohes-

in with chromatin, Cell, 127, 955-967, doi: 10.1016/

j.cell.2006.09.040.

43. Ouyang,Z., Zheng,G., Song,J., Borek, D.M., Otwinows-

ki,Z., Brautigam, C.A., Tomchick, D.R., Rankin,S., and

Yu,H. (2013) Structure of the human cohesin inhibi-

tor Wapl, Proc. Natl. Acad. Sci. USA, 110, 11355-11360,

doi:10.1073/pnas.1304594110.

44. Collier, J.E., Lee, B.-G., Roig, M.B., Yatskevich,S., Pete-

la, N. J., Metson,J., Voulgaris, M., Gonzalez Llamaza-

res,A., Löwe,J., and Nasmyth, K.A. (2020) Transport

of DNA within cohesin involves clamping on top of en-

gaged heads by Scc2 and entrapment within the ring

byScc3, Elife, 9, e59560, doi:10.7554/eLife.59560.

45. Nora, E. P., Caccianini, L., Fudenberg, G., So, K.,

Kameswaran, V., Nagle, A., Uebersohn, A., Hajj, B.,

Saux, A.L., Coulon,A., Mirny, L.A., Pollard, K.S., Da-

han,M., and Bruneau, B.G. (2020) Molecular basis of

CTCF binding polarity in genome folding, Nat. Com-

mun., 11, 5612, doi:10.1038/s41467-020-19283-x.

46. Dauban, L., Montagne, R., Thierry, A., Lazar-Stefan-

ita, L., Bastié, N., Gadal, O., Cournac, A., Koszul, R.,

and Beckouët, F. (2020) Regulation of cohesin-me-

diated chromosome folding by Eco1 and other

partners, Mol. Cell, 77, 1279-1293.e4, doi: 10.1016/

j.molcel.2020.01.019.

COHESIN COMPLEX STRUCTURE 599

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

47. Muir, K. W., Li, Y., Weis, F., and Panne, D. (2020)

The structure of the cohesin ATPase elucidates the

mechanism of SMC-kleisin ring opening, Nat. Struct.

Mol. Biol., 27, 233-239, doi:10.1038/s41594-020-0379-7.

48. Chapard, C., Jones, R., van Oepen, T., Scheinost, J. C.,

and Nasmyth, K. (2019) Sister DNA entrapment be-

tween Juxtaposed Smc heads and Kleisin of the co-

hesin complex, Mol. Cell, 75, 224-237.e5, doi:10.1016/

j.molcel.2019.05.023.

49. Higashi, T. L., Eickhoff, P., Sousa, J. S., Locke, J.,

Nans,A., Flynn, H.R., Snijders, A.P., Papageorgiou,G.,

O’Reilly,N., Chen, Z.A., O’Reilly, F.J., Rappsilber,J., Cos-

ta,A., and Uhlmann,F. (2020) Astructure-based mech-

anism for DNA entry into the cohesin ring, Mol. Cell,

79, 917-933.e9, doi:10.1016/j.molcel.2020.07.013.

50. Bauer, B. W., Davidson, I. F., Canena, D., Wutz, G.,

Tang, W., Litos, G., Horn, S., Hinterdorfer, P., and Pe-

ters, J.-M. (2021) Cohesin mediates DNA loop extrusion

by a “swing and clamp” mechanism, Cell, 184, 5448-

5464.e22, doi:10.1016/j.cell.2021.09.016.

51. Hons, M.T., Huis In ’t Veld, P.J., Kaesler,J., Rombaut,P.,

Schleiffer, A., Herzog, F., Stark, H., and Peters, J.-M.

(2016) Topology and structure of an engineered hu-

man cohesin complex bound to Pds5B, Nat. Commun.,

7, 12523, doi:10.1038/ncomms12523.

52. Ryu, J.-K., Katan, A.J., van der Sluis, E.O., Wisse,T., de

Groot,R., Haering, C.H., and Dekker,C. (2020) Thecon-

densin holocomplex cycles dynamically between open

and collapsed states, Nat. Struct. Mol. Biol., 27, 1134-

1141, doi:10.1038/s41594-020-0508-3.

53. Higashi, T.L., Pobegalov,G., Tang,M., Molodtsov, M.I.,

and Uhlmann, F. (2021) A Brownian ratchet model

for DNA loop extrusion by the cohesin complex, Elife,

10, e67530, doi:10.7554/eLife.67530.

54. Huang, C.E., Milutinovich,M., and Koshland,D. (2005)

Rings, bracelet or snaps: fashionable alternatives for

Smc complexes, Philos. Trans. R. Soc. Lond. B Biol. Sci.,

360, 537-542, doi:10.1098/rstb.2004.1609.

55. Anderson, D.E., Losada,A., Erickson, H.P., and Hira-

no,T. (2002) Condensin and cohesin display different

arm conformations with characteristic hinge angles,

J.Cell Biol., 156, 419-424, doi:10.1083/jcb.200111002.

56. Ciosk, R., Shirayama, M., Shevchenko, A., Tanaka, T.,

Toth, A., Shevchenko,A., and Nasmyth, K. (2000) Co-

hesin’s binding to chromosomes depends on a sepa-

rate complex consisting of Scc2 and Scc4 proteins, Mol.

Cell, 5, 243-254, doi:10.1016/s1097-2765(00)80420-7.

57. Gerlich,D., Koch,B., Dupeux,F., Peters, J.-M., and El-

lenberg,J. (2006) Live-cell imaging reveals a stable co-

hesin-chromatin interaction after but not before DNA

replication, Curr. Biol., 16, 1571-1578, doi: 10.1016/

j.cub.2006.06.068.

58. Uhlmann,F., Lottspeich,F., and Nasmyth,K. (1999) Sis-

ter-chromatid separation at anaphase onset is promot-

ed by cleavage of the cohesin subunit Scc1, Nature,

400, 37-42, doi:10.1038/21831.

59. Minamino, M., Higashi, T. L., Bouchoux, C., and Uhl-

mann,F. (2018) Topological invitro loading of the bud-

ding yeast cohesin ring onto DNA, Life Sci. Alliance, 1,

e201800143, doi:10.26508/lsa.201800143.

60. Haering, C. H., Farcas, A.-M., Arumugam, P., Met-

son,J., and Nasmyth,K. (2008) Thecohesin ring con-

catenates sister DNA molecules, Nature, 454, 297-301,

doi:10.1038/nature07098.

61. Wilhelm,L., Bürmann,F., Minnen,A., Shin, H.-C., Tose-

land, C. P., Oh, B.-H., and Gruber,S. (2015) SMC con-

densin entraps chromosomal DNA by an ATP hydroly-

sis dependent loading mechanism in Bacillus subtilis,

Elife, 4, e06659, doi:10.7554/eLife.06659.

62. Shaltiel, I. A., Datta, S., Lecomte, L., Hassler, M.,

Kschonsak,M., Bravo,S., Stober,C., Ormanns,J., Eus-

termann,S., and Haering, C.H. (2022) Ahold-and-feed

mechanism drives directional DNA loop extrusion

by condensin, Science, 376, 1087-1094, doi: 10.1126/

science.abm4012.

63. Chan, K.-L., Roig, M. B., Hu, B., Beckouët, F., Met-

son, J., and Nasmyth, K. (2012) Cohesin’s DNA exit

gate is distinct from its entrance gate and is regulat-

ed by acetylation, Cell, 150, 961-974, doi:10.1016/j.cell.

2012.07.028.

64. Huis in ’t Veld, P.J., Herzog,F., Ladurner,R., Davidson,

I. F., Piric, S., Kreidl, E., Bhaskara, V., Aebersold, R.,

and Peters, J.-M. (2014) Characterization of a DNA exit

gate in the human cohesin ring, Science, 346, 968-972,

doi:10.1126/science.1256904.

65. Gandhi, R., Gillespie, P. J., and Hirano, T. (2006) Hu-

man Wapl is a cohesin-binding protein that promotes

sister-chromatid resolution in mitotic prophase, Curr.

Biol., 16, 2406-2417, doi:10.1016/j.cub.2006.10.061.

66. Ouyang, Z., and Yu, H. (2017) Releasing the cohes-

in ring: a rigid scaffold model for opening the DNA

exit gate by Pds5 and Wapl, Bioessays, 39, 1600207,

doi:10.1002/bies.201600207.

67. Hauf,S., Waizenegger, I.C., and Peters, J.M. (2001) Co-

hesin cleavage by separase required for anaphase and

cytokinesis in human cells, Science, 293, 1320-1323,

doi:10.1126/science.1061376.

68. Murayama,Y., Samora, C.P., Kurokawa,Y., Iwasaki,H.,

and Uhlmann,F. (2018) Establishment of DNA-DNA in-

teractions by the cohesin ring, Cell, 172, 465-477.e15,

doi:10.1016/j.cell.2017.12.021.

69. Chiu,A., Revenkova,E., and Jessberger,R. (2004) DNA

interaction and dimerization of eukaryotic SMC hinge

domains, J.Biol. Chem., 279, 26233-26242, doi:10.1074/

jbc.M402439200.

70. Soh, Y.-M., Bürmann,F., Shin, H.-C., Oda,T., Jin, K. S.,

Toseland, C.P., Kim,C., Lee,H., Kim, S.J., Kong, M.-S.,

Durand-Diebold, M.-L., Kim, Y.-G., Kim, H. M., Lee,

N. K., Sato, M., Oh, B.-H., and Gruber, S. (2015) Mo-

lecular basis for SMC rod formation and its disso-

lution upon DNA binding, Mol. Cell, 57, 290-303,

doi:10.1016/j.molcel.2014.11.023.

GOLOV, GAVRILOV600

BIOCHEMISTRY (Moscow) Vol. 89 No. 4 2024

71. Griese, J.J., Witte,G., and Hopfner, K.-P. (2010) Struc-

ture and DNA binding activity of the mouse condensin

hinge domain highlight common and diverse features

of SMC proteins, Nucleic Acids Res., 38, 3454-3465,

doi:10.1093/nar/gkq038.

72. Li, Y., Muir, K. W., Bowler, M. W., Metz, J., Haering,

C.H., and Panne,D. (2018) Structural basis for Scc3-de-

pendent cohesin recruitment to chromatin, Elife, 7,

e38356, doi:10.7554/eLife.38356.

73. Kschonsak,M., Merkel,F., Bisht,S., Metz,J., Rybin,V.,

Hassler,M., and Haering, C.H. (2017) Structural basis

for a safety-belt mechanism that anchors condensin

to chromosomes, Cell, 171, 588-600.e24, doi: 10.1016/

j.cell.2017.09.008.

74. Lee, B.-G., Rhodes, J., and Löwe, J. (2022) Clamp-

ing of DNA shuts the condensin neck gate, Proc.

Natl. Acad. Sci. USA, 119, e2120006119, doi: 10.1073/

pnas.2120006119.

Publisher’s Note. Pleiades Publishing remains

neutral with regard to jurisdictional claims in pub-

lished maps and institutional affiliations.