Article

ISSN 0006-2979, Biochemistry (Moscow), 2025, Vol. 90, No. 6, pp. 818-827 © Pleiades Publishing, Ltd., 2025.

Russian Text © The Author(s), 2025, published in Biokhimiya, 2025, Vol. 90, No. 6, pp. 884-895.

818

Designing a Thermostable Mini-Intein

for Intein-Mediated Purification

of Recombinant Proteins and Peptides

Andrey A. Karanov

1,2,a

*, Evgeniy A. Zayats

, Maria A. Kostromina

Yulia A. Abramchik

, Aleksandra R. Sharafutdinova

, Maria S. Surkova

Andrey A. Zamyatnin, Jr.

, and Roman S. Esipov

1,2,b

Institute of Bioorganic Chemistry, Russian Academy of Sciences, 117997 Moscow, Russia

Faculty of Bioengineering and Bioinformatics, Lomonosov Moscow State University,

119234 Moscow, Russia

e-mail: andrey-karanov2000@mail.ru

e-mail: refolding@mail.ru

Received February 7, 2025

Revised May 23, 2025

Accepted May 23, 2025

Abstract— This paper reports the design of a thermostable temperature-activated mini-intein based on the

full-length intein DnaE1 from Thermus thermophilus HB27 (TthDnaE1). We performed rational design of

three mini-inteins TthDnaE1 Δ272, Δ280, and Δ287 through deletion mutations in the full-length intein se-

quence. Two mini-inteins (Δ272 and Δ280) were capable of efficient protein splicing at temperatures above

50°C. The most active mini-intein with the Δ280 deletion was selected as a platform for further design of a

self-cleaving carrier of affinity tags through single-point mutagenesis. Three mutations – C1A, D405G, and

the combined C1A/D405G – were introduced to inhibit N-terminal extein cleavage and extein ligation. As a

result, the mini-intein Δ280 with double mutation C1A/D405G displayed the highest efficiency of C-terminal

extein cleavage with temperature optimum around 60°C. Thus, we constructed thermostable temperature-ac-

tivated mini-intein capable of efficient protein splicing or cleavage of the C-terminal extein. The engineered

TthDnaE1 Δ280 C1A/D405G mini-intein can serve as a basis for the development of new expression system

for intein-mediated production of pharmaceutically relevant recombinant proteins and peptides.

DOI: 10.1134/S0006297925600358

Keywords: recombinant proteins and peptides, mini-inteins, endonuclease domain, thermostability

* To whom correspondence should be addressed.

INTRODUCTION

Development of highly efficient methods for

producing pharmaceutically relevant recombinant

peptides and proteins is a key challenge of medical

biotechnology. Production of these compounds in

prokaryotic expression system meets several difficul-

ties. First, translation in prokaryotes introduces an

N-terminal formylmethionine residue, which often

needs to be removed to obtain a biologically active

molecule [1]. Second, the target product may under-

go intracellular proteolysis in vivo. This phenomenon

is often observed during peptide production [2].

One of the solutions to overcome these issues in-

volves the production of target peptides as a structur-

al part of hybrid proteins followed by their release

through cleavage [3]. Conventionally cleavage of hy-

brid proteins is carried out using site-specific prote-

ases such as TEV protease, SUMO protease, enteroki-

nase, or factor Xa. However, this approach has some

drawbacks: the need to use costly enzymes, limited

efficiency of cleavage, laborious optimization of the

proteolysis conditions, and the need to separate the

unnecessary part of the cleaved hybrid protein [3].

One of the promising alternatives is the con-

struction of self-cleaving hybrid proteins based on

inteins [4, 5]. In this approach the phenomenon of

protein splicing is utilized. Protein splicing is a post

DESIGN OF THERMOSTABLE MINI-INTEIN 819

BIOCHEMISTRY (Moscow) Vol. 90 No. 6 2025

translational process involving autocatalytic exci-

sion of an internal part of the precursor protein

(intein) with simultaneous ligation of the flanking

regions (exteins) [6]. Point mutations may block ex-

tein cleavage and ligation resulting in mutant forms

of inteins that function as autocatalytically cleaving

proteins carrying affinity tags. This approach proved

beneficial. First, it allows to simplify the isolation and

purification of the hybrid protein. Second, it reduces

costs by eliminating the need for using proteases.

Despite these advantages, intein-mediated iso-

lation methods face limitations. A common issue is

premature intracellular cleavage of the hybrid protein

[7-9]. There is also a risk that the efficiency of hy-

brid protein cleavage may be insufficient [10]. Isoelec-

tric point of a hybrid protein may coincide with the

pH range optimal for intein activity thus initiating

irreversible aggregation [11]. To address these lim-

itations, it could be useful to expand the repertoire

of expression constructs based on new inteins.

Conditionally-controlled protein splicing/cleavage

of exteins has been a focal point in recent intein re-

search. Of special interest in this respect are those

inteins that are thermoactivated and thermostable.

Enzymatic activity of these inteins could be con-

trolled by temperature. This approach includes the

synthesis of an inactive intein-containing fusion pro-

tein and further enzyme activation via structure res-

toration. For instance, Shen et  al. used thermostable

and thermoactivated intein to regulate the activity of

a xylanase for lignocellulose biomass processing [12].

Wang etal. employed a thermoregulated intein to con-

trol the activity of a thermostable DNA polymerase

from Pseudomonas fluorescens, potentially enhancing

PCR techniques [13].

In this study we suggest a new application of ther-

mal activation and thermostability in inteins for use

in purification of recombinant proteins and peptides.

High activation temperature of intein may prevent

premature invivo cleavage. Furthermore, thermal sta-

bility enables initial purification via temperature-me-

diated precipitation of ballast proteins.

It is necessary to consider the domain organisa-

tion of inteins during the development of intein-based

biotechnological tools. Inteins are generally classi-

fied as either full-length or mini-inteins. Full-length

inteins consist of two domains. The HINT-domain

(Hedgehog-Intein) responsible for protein splicing

is formed by the intein's N- and C-terminal regions

(designated HgN and HgC). The second domain with

endonuclease activity is located in the middle part

of intein, thus splitting the HINT-domain in two

parts. The structure of mini-inteins includes only the

HINT-domain, while the endonuclease domain is ab-

sent. Mini-inteins can be either natural or artificial.

Small size of mini-inteins is advantageous, as well as

lack of endonuclease domain, which may negative-

ly affect solubility of the hybrid protein. Artificial

mini-inteins derived from the DnaB of Synechocys-

tis sp. and RecA of Mycobacterium tuberculosis have

been shown to be the most useful tools for isolation

and purification of proteins and peptides [14-19].

This study aims at the development of expression

system for isolation of pharmaceutically relevant pep-

tides and proteins based on the thermoactivated and

thermostable DnaE1 intein of Thermus thermophilus

(TthDnaE1). Shen etal. have shown that TthDnaE1 is a

thermostable full-length intein capable of performing

protein splicing at 60°C [12]. The goal of this study is

the rational design of modified mini-inteins based on

TthDnaE1 through deletion of endonuclease domain

and introducing point mutations at catalytic residues,

as well as evaluation of their practical applicability.

MATERIALS AND METHODS

Generation of strains producing recombinant

mini-inteins. To construct genes of artificial mini-in-

teins, two fragments of the dnaE gene were PCR-am-

plified using chromosomal DNA of Thermus ther-

mophilus strain HB27 as a template. One fragment

encoded N-terminal (upstream of the deletion) region

of the intein. The other fragment encoded C-terminal

sequence downstream of the deletion. Primers for the

N-terminal fragment included recognition sites for

the restriction endonucleases NdeI and Esp3I, while

primers for the C-terminal fragment included Esp3I

and Bpu1102I sites. Amplification was carried out us-

ing Q5 High-Fidelity DNA Polymerase (New England

Biolabs, USA). Products of amplification were digested

with Esp3I restriction endonuclease and ligated with

the help of a bacteriophage T4 ligase (Thermo Fisher

Scientific, USA). The ligated fragment was amplified

and cloned into a pET16b vector at recognition sites

of the restriction endonucleases NdeI and Bpu1102I.

The resulting sequences of mini-inteins were verified

by Sanger sequencing (Evrogen, Russia).

Producer strains were obtained by transforming

engineered plasmids into E.  coli NiCo21 (DE3) cells

(New England Biolabs, USA). Cultivation was per-

formed in LB medium (10  g/l trypton, 5  g/l yeast ex-

tract, 10 g/l NaCl) supplemented with 100  µg/ml am-

picillin. Cell cultures were grown in a thermostated

shaker Certomat SII (Sartorius Stedim Biotech, Germa-

ny) at 37°C and 180 rpm until optical density OD

600

0.8

followed by addition of IPTG to final concentration

0.4  mM and further incubation for 4  h at 37°C. All tar-

get proteins were expressed in a soluble form.

Isolation and purification of recombinant mini-

inteins. Cell biomass was resuspended at a ratio

1  :  10 (w/v) in buffer  A (50  mM  Tris-HCl, 200  mM  NaCl,

KARANOV et al.820

BIOCHEMISTRY (Moscow) Vol. 90 No. 6 2025

5  mM  EDTA (Merck Millipore, Germany), 1  mM  PMSF

(Sigma-Aldrich, USA), pH 8.0) and disrupted using an

ultrasonic disintegrator VCX  130 (Sonics Materials,

USA). Ammonium sulfate (Panreac, Spain) was added

to the cell supernatant to 20% saturation, followed by

centrifugation for 20  min at 10,000g. Precipitate con-

taining ballast proteins was discarded. To precipitate

the target proteins, ammonium sulfate was added to

the prepared supernatant to 40% saturation. The re-

sulting pellet was resuspended in buffer  B (50  mM

Tris-HCl, 200  mM  NaCl, pH  8.0) and clarified by cen-

trifugation. The clarified solution was loaded onto

a XK16/20 column packed with Chelating Sepharose

Fast Flow resin (Cytiva, USA). The column was washed

with buffer  C (50  mM  Tris-HCl, 200  mM  NaCl, 50  mM

imidazole, pH  8.0) to elute non-target proteins. In the

second stage a target protein was eluted with buffer  D

(50  mM  Tris-HCl, 200  mM  NaCl, 250  mM  imidazole,

pH  8.0). EDTA was added to the eluted fraction to fi-

nal concentration 5mM.

Fractions containing the target protein were

pooled and concentrated to 5 ±  1  mg/ml using a YM-10

membrane (Merck Millipore). The final purification

step involved gel-filtration chromatography on a

HiLoad 16/60 Superdex 75  pg column (Cytiva) in buf-

fer  E (1  mM  NaH

, 20  mM  Na

HPO

, 200  mM NaCl,

5  mM  EDTA, pH 8.0). Fractions containing a target pro-

tein were combined.

Evaluation of mini-intein protein splicing effi-

ciency. Efficiency of protein splicing was investigated

in a pH range from 6.0 to 9.0 at 60°C. For this pur-

pose, purified mini-intein samples were diluted with

buffer E to final concentration of 0.3 mg/ml, and pH

was adjusted. Temperature dependence of splicing

was evaluated at pH 6.0 after 14-h incubation in the

temperature range from 20 to 80°C. Splicing efficiency

was assessed by electrophoretic analysis.

Analytical methods. Protein concentrations were

determined using the Bradford protein assay [20].

Electrophoretic analysis was carried out according to

the Laemmli technique in a 15%  SDS-PAGE [21]. Den-

sitometric analysis of gels was performed in three

replicates to calculate average splicing efficiencies.

To obtain electropherograms with enhanced resolu-

tion in the peptide range of molecular mass, a Tris-tri-

cine buffer system was used for electrophoresis in

a  10%  SDS-PAGE [22].

RESULTS

Rational design of thermostable mini-inteins.

The key point in rational design of mini-inteins is

identification of a boundary between the endonucle-

ase domain and HINT-domain in the full-length intein.

The structure of TthDnaE1 [including natural exteins

each containing 5 amino acid residues (aa)] was

modeled using AlphaFold  3. Superposition of the ob-

tained model with the structure of the artificial mini-

intein SspDnaB (PDB  ID: 1MI8) enabled identification

Fig. 1. Comparison of the structures of TthDnaE1 and artificial mini-intein SspDnaB. N- and C-terminal regions of the poly-

peptide chain forming HINT-domain are shown in green and orange, respectively. β-Strands formed by amino acid residues

G87-V92 and L388-L406 are shown in blue and red, respectively.

DESIGN OF THERMOSTABLE MINI-INTEIN 821

BIOCHEMISTRY (Moscow) Vol. 90 No. 6 2025

Fig. 2. Structure models of mini-inteins TthDnaE1 Δ272 (a), Δ280 (b), and Δ287 (c).

oftheHINT-domain in TthDnaE1 (Fig.1). We illustrat-

ed the similarity between the structure of the HINT-

domain in the TthDnaE1 model and crystal structure

of SspDnaB: root mean square deviation of the co-

ordinates of carbon atoms in the peptide backbone

was around 0.7  Å. β-Strands formed by the amino

acid residues G87-V92 and L388-L406 were the struc-

tural elements closest to the TthDnaE1 endonuclease

domain.

First, we analyzed the previously reported ex-

amples of rational design of mini-inteins. There are

at least 11 examples of full-length inteins for which

removal of the endonuclease domain without loss

of splicing activity was shown: MtuRecA, SspDnaB,

NpuDnaB, RmaDnaB, SspDnaX, SspGyrB, TerThyX,

TerRIR1, SceVma, Pi-PFUI, and PhoVMA [15, 23-26].

One intein of particular interest was DnaE2 of Tri-

chodesmium erythraeum (TerDnaE2). Removal of the

endonuclease domain from this intein resulted in

the significant decrease in splicing efficiency [25,  27].

Nevertheless, this intein is interesting since it is an

ortholog of TthDnaE1.

Amino acid sequences of the TthDnaE1 and of the

abovementioned inteins were compared using multiple

sequence alignment with the Clustal Omega (Fig.S1 in

the Online Resource 1) [28]. The obtained alignment

was used to generate a phylogenetic tree (Fig.S2 in in

the Online Resource 1) using Mega program (UPGMA

method) [29]. The inteins DnaB of Synechocystis sp.

(SspDnaB), Nostoc punctiforme (NpuDnaB), and Rhodo-

thermus marinus (RmaDnaB) were shown to be the

closest relatives to the investigated intein. Hence, in

order to select deletions, it is most feasible to compare

the intein investigated in our study with these inteins.

It should be also mentioned that among those the

SspDnaB intein is especially notable due to its wide

application for intein-mediated isolation of proteins

and peptides [6, 14-17].

We decided to select three deletions, which

could theoretically result in formation of loops by

the remaining residues of endonuclease domain with

length difference of approximately 8  aa. We consid-

ered deletion variants that are homologous to those

that resulted in active artificial mini-inteins derived

from DnaB. Three deletions in TthDnaE1 correspond

to this criterium, which are termed according to the

number of deleted residues: Δ272, Δ280, and Δ287.

The mini-intein TthDnaE1 obtained as a result of

the Δ272 deletion includes N- and C-terminal frag-

ments of the amino acid sequence homologous to

the fragments of polypeptide chain of the SspDnaB

mini-intein (Table  S1 in in the Online Resource1). The

variants containing deletions Δ280 and Δ287 include

the fragments homologous to the N- and C-terminal

fragments of the two artificial mini-inteins NpuDnaB

(Table  S1 in the Online Resource  1). The structures of

the suggested mini-inteins [with natural exteins each

containing 5 amino acid residues (aa)] were gener-

ated with AlphaFold  3.0  [30]. It was shown that the

structure of HINT-domain in all three mini-inteins is

identical, while the remaining residues of the endonu-

clease domain form disordered loop HgN-X-HgC with

size of 23, 15, and 8 aa in the TthDnaE1 Δ272, Δ280,

and Δ287, respectively (Fig.2; see also domain scheme

in Fig. S3 in the Online Resource 1).

Structural stability of the HINT-domain of the

artificial mini-inteins was investigated in silico

through molecular dynamics simulations (MD) with

Gromacs  2024 and the Amber ff99SB-ILDN force field

(detailed description of the used MD methods is pre-

sented in the Online Resource 1) [31,  32]. Each mod-

el was placed into rhombic dodecahedron unit cells

filled with tip3p water molecules and Na

and Cl

−

ions

to neutralize charge. After equilibration at 60°C and

pressure 1  Bar by modeling in the NVT- and NPT-en-

sembles, productive MD simulation with duration of

500 ns was performed.

To examine stability of the structures of the mini

inteins during MD simulation we calculated Root

Mean Square Deviation (RMSD), Root Mean Square

Fluctuation (RMSF) of the atomic positions of the

peptide backbone and the radius of gyration of the

protein molecules. It should be mentioned that struc-

ture instability is expected for the exteins and loops

KARANOV et al.822

BIOCHEMISTRY (Moscow) Vol. 90 No. 6 2025

connecting HgN and HgC. In this regard, the RMSD

values and radius of gyration were calculated not

only for the entire protein molecule, but also for the

HINT-domain separately. The obtained results are

presented in Fig. S4 in the Online Resource 1.

Analysis of the RMSD values and radius of gyra-

tion indicated high stability of the HINT-domain in

all three mini-inteins in the course of MD simulation.

Theprofile of RMSF values indicates that exteins and

HgN-X-HgC loop are the sites of instability in the poly-

peptide chain. Furthermore, the HgN-X-HgC loop in

the mini-intein Δ280 was significantly more dynamic

in comparison with two other variants.

One other factor that should be considered in the

design of mini-inteins is the effect of the HgN-X-HgC

loop on the protein solubility. This loop can include

fragments of the polypeptide chain that forms the core

of the endonuclease domain globule in the full-length

intein. In this case hydrophobicity of amino acid res-

idues of the unordered loop could negatively affect

the protein solubility.

Therefore, we analyzed the potential effects of the

sequence and spatial structure of the designed mini-

inteins on solubility of the proteins using CamSol  2.0

web server [33]. The approach of the prediction of

such effects of the fragments of polypeptide chain is

based on: 1)  analysis of amino acid sequence based

on physicochemical properties of amino acid residues;

2)  correction of the obtained results considering pro-

tein spatial structure. Structure models of the sug-

gested mini-inteins were analyzed using CamSol  2.0.

It was shown that the loops HgN-X-HgC do not contain

amino acid residues which could affect solubility neg-

atively (Fig. S5 in the Online Resource 1, a-c).

Besides the design of mini-inteins, another im-

portant challenge is the selection of the lengths of

the extein fragments extein for further investigation

of protein splicing in vitro. Intein activity could be

significantly affected by the amino acid sequence of

an extein. That is why it is reasonable to investigate

protein splicing with natural exteins. On the other

hand, use of inteins for protein purification is espe-

cially relevant in the case of peptides with unordered

structure. Therefore, we decided to investigate protein

splicing by using hybrid constructs containing mini-in-

teins with fragments of natural exteins up to 20  aa

in length.

Certain fragments of natural exteins could par-

ticipate in formation of the globule of natural DnaE

protein. Presence of hydrophobic sites may negatively

affect solubility. To resolve this issue the size of extein

fragments was optimized with the help CamSol 2.0 web

server. Structure model of the full-length intein with

fragments of natural exteins with length of 20  aa was

obtained by using of neural network AlphaFold  3.0.

Analysis of the obtained model with the help of

CamSol2.0 showed that the exteins larger than 10  aa

at the N-end could significantly reduce the protein sol-

ubility (Fig. S5 in the Online Resource 1, d).

The final step of the design was selection of point

mutations in order to change the intein capability of

protein splicing. The sequence of TthDnaE1 was com-

pared with the sequence of MtuRecA intein, for which

key amino acids participating in protein splicing have

been determined (based on the previously performed

alignment of amino acid sequences)[23]. This allowed

to identify homologous residues in TthDnaE1: Cys1,

Asp405, His422, and Asn423. Positions of these resi-

dues in the model structure of TthDnaE1 Δ280 intein

are shown in Fig. S6 in in the Online Resource 1.

Protein splicing occurs due to rearrangement with

participation of hydroxyl/thiol groups of the first ami-

no acid of intein and the first amino acid of the C-end

of extein (in the case of TthDnaE1 these are Cys1 and

Ser+1). The Cys1Ala mutation disables cleavage of the

N-terminal extein. It was shown that the highly con-

served Asp residue is significant for protein splicing

(Asp405 in TthDnaE1) [23]. Substitution of this Asp

residue with Gly was shown to significantly accelerate

the cleavage of C-terminal extein. Hence, the D405G

mutation is practical for the TthDnaE1 both as a sin-

gle mutation and in combination with C1A. Two last

residues in the intein (His422, Asn423 in TthDnaE1)

participate in the cleavage of the C-terminal extein.

Mutations in these positions are not practical because

cleavage of the C-terminal extein is most significant

for the development of the approach to intein-medi-

ated isolation and purification of proteins.

Hence, modeling of the protein spatial struc-

ture with the help AlphaFold 3.0, MD simulations in

Gromacs, and predicting effects of amino acid on the

protein solubility allowed us to perform rational de-

sign of three mini-inteins: TthDnaE1 Δ272, Δ280, and

Δ287. Analysis of functional amino acids in the intein

allowed us to select the C1A, D405G, and C1A/D405G

mutations that prevent cleavage of the N-terminal ex-

tein.

Preparation and investigation of properties

of artificial mini-inteins. Expression constructs en-

coding mini-inteins TthDnaE1 Δ272, Δ280, and Δ287

flanked with the fragments of natural exteins (N-ex-

tein with length 10  aa and C-extein with length 16  aa)

and N-terminal affinity tag 6x-His were created, and

the corresponding E.  coli producer strains were gen-

erated. Recombinant proteins were isolated using a

two-stage protocol including purification via affinity

and gel-filtration chromatography.

Temperature and pH dependence of splicing effi-

ciency (range 20-80°C at pH 6.0; range 6.0-9.0 at 60°C,

respectively) was evaluated based on the results of

incubation of the reaction mixtures for 14  h. Four

possible products were expected (in accordance with

DESIGN OF THERMOSTABLE MINI-INTEIN 823

BIOCHEMISTRY (Moscow) Vol. 90 No. 6 2025

Fig. 3. Electrophoretic analysis of the products of splicing

of hybrid proteins containing mini-intein TthDnaE1 with

Δ272, Δ280, and Δ287 deletions (15% SDS-PAAG). Designa-

tions: EIE, non-cleaved mini-intein; EI, residual protein

without C-terminal exteins; IE, residual protein without

N-terminal extein; I, mini-intein without exteins; EE, product

of ligation of exteins.

the calculated molecular mass, from smallest to larg-

est): 1)  product of extein ligation (designated as EE;

6.2  kDa); 2)  mini-intein without exteins (product  I;

16.4  kDa for Δ272, 15.5  kDa for Δ280); 3)  residual pro-

tein without N- terminal extein (product IE; 18.3  kDa

for Δ272, 17.5  kDa for Δ280); 4)  residual protein with-

out C-terminal extein (product EI; 20.5  kDa for Δ272,

19.7  kDa for Δ280).

The results of electrophoretic analysis of the

mini-inteins splicing products after incubation at 60°C

for 14  h at pH  6.0 are presented in Fig.3. Cleavage of

the proteins with deletion Δ272 and Δ280 (EIE bands)

resulted in formation of side products in addition to

the main products (EE and I).

The TthDnaE1 Δ280 intein was found to be the

most active: almost complete cleavage of the corre-

sponding hybrid protein was observed after incuba-

tion. The Δ272 deletion variant exhibited significantly

lower activity, accompanied by high level of proteol-

ysis of both hybrid protein and of the cleavage prod-

ucts observed. The variant Δ287 showed no detectable

splicing activity.

These results confirm the success of the rational

design approach, yielding a mini-intein variant (Δ280)

capable of efficient, temperature-dependent protein

splicing. This construct was therefore selected for fur-

ther development of the self-cleaving carrier protein.

Next, site-directed mutagenesis of the gene of

mini-intein TthDnaE1 Δ280 was performed to intro-

duce point mutations C1A, D405G, and C1A/D405G.

The resulting hybrid proteins were expressed using

the same method as for wild type constructs. Splicing

activity of the modified mini-inteins was examined

according to the abovementioned protocol.

Comparison of the products of cleavage of the

hybrid proteins based on the generated mini-intein

Δ280 and on its mutant variants C1A, D405G, and C1A/

D405G is presented in Fig.4. Introduction of all three

mutations resulted in significant changes in compo-

sition and relative abundance of the formed splicing

products.

Splicing efficiency in all three mutant con-

structs was strongly dependent on both tempera-

ture and pH (Fig.  5). Optimal range of pH (6.0-6.5)

for cleaving was determined for all tested variants.

Fig. 4. Electrophoretic analysis of the products of splicing of hybrid proteins based on the mini-intein TthDnaE1 Δ280

without mutations and with point substitutions C1A, D405G, and C1A/D405G. a) 15% SDS-PAAG; b) 10% tricine SDS-PAAG.

Designations are the same as in Fig. 3; EN, N-terminal extein; EC, C-terminal extein.

KARANOV et al.824

BIOCHEMISTRY (Moscow) Vol. 90 No. 6 2025

Fig. 5. Products of cleavage of the hybrid proteins based on the TthDnaE1 Δ280 mini-intein (a, b) and its mutant forms

C1A(c,d), D405G(e,f), and C1A/D405G(g,h) depending on the incubation conditions. a, c, e, g)Content of products formed

at pH6.0 and temperature in the range 20-80°C; b, d, f, h)at 60°C and pH in the range 6.0-9.0. Initial hybrid protein marked

with white color, product I – with light gray, EI product – with dark gray. The data are obtained using densitometry of the

15% SDS+PAAGs. C, control, hybrid protein before incubation.

Increasing the incubation temperature enhanced effi-

ciency of cleavage significantly: hybrid protein based

on the mini-intein TthDnaE1 Δ280 exhibited 79%

(at 60°C and pH 6.0) versus 27% (at 20°C). Similar

results were obtained for the mutant forms. When

incubated at 70°C and pH 6.0-6.5, the cleavage effi-

ciency increased notably: from 4% to 31% for the

C1A mutant, from 5% to 42% for the D405G mutant,

from 11% to 54% for the C1A/D405G mutant. Optimal

incubation time of 14 h was selected based on the

data of the cleavage kinetics. After 14-h incubation

accumulation of the products slows down; moreover,

prolonged incubation may lead to proteolytic degrada-

tion of the products (Fig.S7 in the Online Resource1).

DESIGN OF THERMOSTABLE MINI-INTEIN 825

BIOCHEMISTRY (Moscow) Vol. 90 No. 6 2025

All expected cleavage products were detected for

protein carrying D405G mutation: cleavage of both ex-

teins and cleavage of C- or N-terminal extein separate-

ly (product EN, 4.2  kDa; formation of the EC product

was detected while increasing the gel load). At the

same time no extein ligation was observed. Cleavage

of the C- and N-terminal exteins was confirmed by

electrophoretic analysis in the Tris-tricine buffer sys-

tem (Fig.  5b). Hence, the D405G mutation significant-

ly increases efficiency of cleavage of the C-terminal

extein and blocks ligation of exteins.

Cleavage of only C-terminal extein (yielding prod-

ucts EI and EC) was observed for C1A and C1A/D405G

variants. Hence, this mutation resulted in the desir-

able effect– prevention of cleavage of the N-terminal

extein. Notably, the hybrid protein with double muta-

tion exhibited the highest cleavage efficiency.

DISCUSSION

In this study, we successfully performed the ra-

tional design of the thermoactivated mini-intein based

on the full-length DnaE1 intein from Thermus ther-

moplilus.

An approach that includes amino acid sequence

analysis, structure modeling with AlphaFold 3.0 and

MD simulations allowed us to select three deletion

variants (Δ272, Δ280, and Δ287) for removal of the en-

donuclease domain from the initial TthDnaE1 intein.

Experimental investigation of the engineered mini-

inteins showed that the medium-sized deletion was

the least disruptive for protein splicing – the TthD-

naE1 Δ280 mini-intein exhibited the highest activity.

The reasons behind the reduced activity of the Δ272

and Δ287 mini-inteins require further investigation.

It should be mentioned that AlphaFold structure mod-

eling and MD simulation in Gromacs are not abinitio

approaches. It is possible that the negative effects of

the Δ272 and Δ287 deletions could be explained by

their effects exactly on the processes of folding of mini-

inteins, which are challenging to predict.

Next, we have investigated the effects of point

mutations at the key positions responsible for pro-

tein splicing. Two targets were selected for mutagen-

esis: Cys1 and Asp405. The ability of cleavage of the

N-terminal extein was blocked by the C1A mutation.

The single D405G mutation resulted in prevention of

ligation of the cleaved exteins. The obtained results

indicate significance of the Cys1 and Asp405 residues

for different stages of protein splicing.

The capability of C-terminal extein cleavage is

valuable for the intein-mediated isolation of proteins

and peptides. All other activities of intein may result

in formation of non-desired side products. The C1A/

D405G mini-intein variant demonstrated the desirable

result: the only product was C-terminal extein while

side products formation was not observed. Hence, it

was shown that the mini-intein TthDnaE1 Δ280 C1A/

D405G is capable of effective cleavage of the C-ter-

minal extein, and constitutes a promising object for

further studies.

It was shown that temperature optimum of the

mini-intein engineered in this study is in the range

of 60-70°C. Therefore, the efficiency of the cleavage of

the C-terminal extein is reduced at lower temperatures

(20-40°C), which corresponds to the standard cultiva-

tion temperatures of the E. coli producer strains. This

feature could be useful to overcome the premature

intracellular cleavage of hybrid proteins.

The design of an expression system based on the

TthDnaE1 Δ280 C1A/D405G mini-intein and demon-

stration of its applicability for production of a num-

ber of model peptides seems to be a logical next step

of the study.

CONCLUSIONS

Development of new approaches for the isolation

and purification of recombinant proteins and pep-

tides is one of the key challenges of biotechnology.

Construction of hybrid proteins that include the target

peptide sequence provides a solution for many techno-

logical problems characteristic to prokaryotic systems

such as presence of N-terminal formylmethionine or

intracellular degradation. Inteins were successfully

proven themselves to be useful as affinity tag carri-

ers and simultaneously autocatalytic proteases. Design

of novel inteins exhibiting new methods of control of

protein splicing and cleavage of exteins helps to solve

the key challenges of this technology.

In this study we engineered a thermostable and

thermoactivated mini-inteins based on the full-length

DnaE1 intein from Thermus thermophilus HB27.

Properties of such inteins have some significant ad-

vantages for their use in intein-mediated protein pu-

rification. High temperature optimum for the cleav-

age of exteins could prevent cleavage of this hybrid

protein in  vivo at temperatures optimal for cultiva-

tion of mesophilic producer strains. In future studies

we plan to develop an expression system using the

TthDnaE1 Δ280 mini-intein with double mutation C1A/

D405G and test this system for the production of phar-

maceutically significant peptides.

Abbreviations. MD, molecular dynamics.

Supplementary information. The online version

contains supplementary material available at https://

doi.org/10.1134/S0006297925600358.

Contributions. R. S. Esipov and A. A. Zamyatin –

concept and supervision of the study; A. A. Karanov,

KARANOV et al.826

BIOCHEMISTRY (Moscow) Vol. 90 No. 6 2025

E. A. Zayats, A. R. Sharafutdinova, and M. S. Surkova–

conducting experiments; A. A. Karanov, E. A. Zayats,

M. A. Kostromina, and Yu. A. Abramchik – discussion

of the results of the study; A. A. Karanov and E. A.

Zayats – writing text of the paper; R. S. Esipov, M. A.

Kostromina, and Yu. A. Abramchik– editing text of the

paper.

Ethics approval and consent to participate.

This work does not contain any studies involving hu-

man and animal subjects.

Conflict of interest. The authors of this work

declare that they have no conflicts of interest.

REFERENCES

1. Ben-Bassat, A. (1991) Purification and Analysis of Re-

combinant Proteins (Seetharam, R., and Sharma, S. K.,

eds), Marcel Dekker, N.Y., pp.147-151.

2. Enfors, S. O. (1992) Control of in vivo proteolysis

in the production of recombinant proteins, Trends

Biotechnol, 10, 310-315, https://doi.org/10.1016/0167-

7799(92)90256-u.

3. Mahmoodi, S., Pourhassan-Moghaddam, M., Wood,

D. W., Majdi, H., and Zarghami, N. (2019) Current

affinity approaches for purification of recombinant

proteins, Cogent Biol., 5, 1665406, https://doi.org/

10.1080/23312025.2019.1665406.

4. Eskandari, A., Leow, T. C., Rahman, M. B. A., and

Oslan, S. N. (2024) Utilization and prospect of purifi-

cation technologies in natural proteins, peptides and

recombinant proteins, J. Proteins Proteom., 15, 233-

257, https://doi.org/10.1007/s42485-024-00139-7.

5. Esipov, R. S., Stepanenko, V. N., Chupova, L. A.,

Boyarskikh, U. A., Filipenko, M. L., and Miroshnikov,

A. I. (2008) Production of recombinant human epi-

dermal growth factor using Ssp dnaB mini-intein

system, Protein Expr. Purif., 61, 1-6, https://doi.org/

10.1016/j.pep.2008.05.009.

6. Volkmann,G., and Mootz, H.D. (2013) Recent progress

in intein research: from mechanism to directed evo-

lution and applications, Cell Mol. Life Sci., 70, 1185-

1206, https://doi.org/10.1007/s00018-012-1120-4.

7. Yuan, H., Prabhala, S. V., Coolbaugh, M. J., Stimple,

S. D., and Wood, D.W. (2024) Improved self-cleaving

precipitation tags for efficient column free biosep-

arations, Protein Expr. Purif., 224, 106578, https://

doi.org/10.1016/j.pep.2024.106578.

8. Lahiry, A., Fan, Y., Stimple, S. D., Raith, M., and

Wood, D. W. (2018) Inteins as tools for tagless and

traceless protein purification, J. Chem. Technol.

Biotechnol., 93, 1827-1835, https://doi.org/10.1002/

jctb.5415.

9. Cui, C., Zhao, W., Chen, J., Wang, J., and Li,Q. (2006)

Elimination of in vivo cleavage between target pro-

tein and intein in the intein-mediated protein purifi-

cation systems, Protein Expr. Purif., 50, 74-81, https://

doi.org/10.1016/j.pep.2006.05.019.

10. Hiraga, K., Derbyshire, V., Dansereau, J. T., Van

Roey,P., and Belfort,M. (2005) Minimization and sta-

bilization of the Mycobacterium tuberculosis recA in-

tein, J.Mol. Biol., 354, 916-926, https://doi.org/10.1016/

j.jmb.2005.09.088.

11. Esipov, R. S., Beirakhova, K. A., Chupova, L. A.,

Likhvantseva, V.G., Stepanova, E.V., and Miroshnokov,

A.I. (2012) Recombinant fragment of pigment epithe-

lium-derived factor (44-77) prevents pathological cor-

neal neovascularization, Bioorg. Khim., 38, 1-8.

12. Shen,B., Sun,X., Zuo,X., Shilling,T., Apgar,J., Ross,M.,

Bougri, O., Samoylov, V., Parker, M., Hancock, E.,

Lucero,H., Gray,B., Ekborg, N.A., Zhang,D., Johnson,

J.C.S., Lazar,G., and Raab, R.M. (2012) Engineering a

thermoregulated intein-modified xylanase into maize

for consolidated lignocellulosic biomass process-

ing, Nat. Biotechnol., 30, 1131-1136, https://doi.org/

10.1038/nbt.2402.

13. Wang, Y., Shi, Y., Hellinga, H. W., and Beese, L. S.

(2023) Thermally controlled intein splicing of en-

gineered DNA polymerases provides a robust and

generalizable solution for accurate and sensitive mo-

lecular diagnostics, Nucleic Acids Res., 51, 5883-5894,

https://doi.org/10.1093/nar/gkad368.

14. Yan, S.-S., Yan,J., Shi,G., Xu,Q., Chen, S.-C., and Tian,

Y.-W. (2005) Production of native protein by using

Synechocystis sp. PCC6803 DnaB mini-intein in Esch-

erichia coli, Protein Expr. Purif., 40, 340-345, https://

doi.org/10.1016/j.pep.2004.12.021.

15. Wu, H., Xu, M.-Q., and Liu, X.-Q. (1998) Protein

trans- splicing and functional mini-inteins of a cy-

anobacterial dnaB intein, Biochim. Biophys. Acta,

1387, 422-432, https://doi.org/10.1016/s0167-4838

(98)00157-5.

16. Malone, C. L., Boles, B. R., and Horswill, A. R. (2007)

Biosynthesis of Staphylococcus aureus autoinducing

peptides by using the Synechocystis DnaB mini-in-

tein, Appl. Environ. Microbiol., 73, 6036-6044, https://

doi.org/10.1128/AEM.00912-07.

17. Tian, L., and Sun, S. S. (2011) A cost-effective

ELP-intein coupling system for recombinant pro-

tein purification from plant production platform,

PLoS One, 6, e24183, https://doi.org/10.1371/journal.

pone.0024183.

18. Gangopadhhyay, J. P., Jiang, S.-Q., van Berkel, P., and

Paulus, H. (2003) In vitro splicing of erythropoie-

tin by the Mycobacterium tuberculosis RecA intein

without substituting amino acids at the splice junc-

tions, Biochim. Biophys. Acta, 1619, 193-200, https://

doi.org/10.1016/s0304-4165(02)00495-6.

19. Banki, M. R., Feng, L., and Wood, D. W. (2005) Sim-

ple bioseparations using self-cleaving elastin-like

polypeptide tags, Nat. Methods, 2, 659-661, https://

doi.org/10.1038/nmeth787.

DESIGN OF THERMOSTABLE MINI-INTEIN 827

BIOCHEMISTRY (Moscow) Vol. 90 No. 6 2025

20. Bradford, M. M. (1976) A rapid and sensitive meth-

od for the quantitation of microgram quantities of

protein utilizing the principle of protein-dye binding,

Anal. Biochem., 72, 248-254, https://doi.org/10.1016/

0003-2697(76)90527-3.

21. Laemmli, U.K. (1970) Cleavage of structural proteins

during the assembly of the head of bacteriophage

T4, Nature, 227, 680-685, https://doi.org/10.1038/

227680a0.

22. Schägger,H., and von Jagow,G. (1987) Tricine- sodium

dodecyl sulfate-polyacrylamide gel electrophoresis

for the separation of proteins in the range from 1

to 100 kDa, Anal. Biochem., 166, 368-379, https://

doi.org/10.1016/0003-2697(87)90587-2.

23. Van Roey,P., Pereira,B., Li,Z., Hiraga,K., Belfort,M.,

and Derbyshire, V. (2007) Crystallographic and mu-

tational studies of Mycobacterium tuberculosis recA

mini-inteins suggest a pivotal role for a highly con-

served aspartate residue, J. Mol. Biol., 367, 162-173,

https://doi.org/10.1016/j.jmb.2006.12.050.

24. Aranko, A. S., Oeemig, J. S., Zhou, D., Kajander, T.,

Wlodawer,A., and Iwaï,H. (2014) Structure-based en-

gineering and comparison of novel split inteins for

protein ligation, Mol. Biosyst., 10, 1023-1034, https://

doi.org/10.1039/c4mb00021h.

25. Lin, Y., Li, M., Song, H., Xu, L., Meng, Q., and Liu,

X. Q. (2013) Protein trans-splicing of multiple atypi-

cal split inteins engineered from natural inteins, PLoS

One, 8, e59516, https://doi.org/10.1371/journal.pone.

0059516.

26. Chong, S., Mersha, F. B., Comb, D. G., Scott, M. E.,

Landry, D., Vence, L. M., Perler, F. B., Benner, J.,

Kucera, R.B., Hirvonen, C.A., Pelletier, J.J., Paulus,H.,

and Xu, M. (1997) Single-column purification of free

recombinant proteins using a self-cleavable affin-

ity tag derived from a protein splicing element,

Gene, 192, 271-281, https://doi.org/10.1016/s0378-1119

(97)00105-4.

27. Perler, F. B. (2002) InBase, the intein database, Nu-

cleic Acids Res., 30, 383-384, https://doi.org/10.1093/

nar/30.1.383.

28. Sievers, F., Wilm, A., Dineen, D., Gibson, T. J.,

Karplus, K., Li, W., Lopez, R., McWilliam, H.,

Remmert,M., Söding,J., Thompson, J.D., and Higgins,

D. G. (2011) Fast, scalable generation of high-quality

protein multiple sequence alignments using Clustal

Omega, Mol. Syst. Biol., 7, 539, https://doi.org/10.1038/

msb.2011.75.

29. Hall, B. G. (2013) Building phylogenetic trees from

molecular data with MEGA, Mol. Biol. Evol., 30, 1229-

1235, https://doi.org/10.1093/molbev/mst012.

30. Abramson,J., Adler,J., Dunger,J., Evans,R., Green,T.,

Pritzel,A., Ronneberger,O., Willmore,L., Ballard, A.J.,

Bambrick, J., Bodenstein, S. W., Evans, D. A., Hung,

C. C., O’Neill, M., Reiman, D., Tunyasuvunakool, K.,

et al. (2024) Accurate structure prediction of biomo-

lecular interactions with AlphaFold 3, Nature, 630,

493-500, https://doi.org/10.1038/s41586-024-07487-w.

31. Van der Spoel, D., Lindahl, E., Hess, B., Groenhof, G.,

Mark, A. E., and Berendsen, H. J. (2005) GROMACS:

fast, flexible, and free, J.Comput. Chem., 26, 1701-

1718, https://doi.org/10.1002/jcc.20291.

32. Lindorff-Larsen,K., Piana,S., Palmo,K., Maragakis,P.,

Klepeis, J. L., Dror, R. O., and Shaw, D. E. (2010) Im-

proved side-chain torsion potentials for the Amber

ff99SB protein force field, Proteins, 78, 1950-1958,

https://doi.org/10.1002/prot.22711.

33. Sormanni,P., Aprile, F.A., and Vendruscolo,M. (2015)

The CamSol method of rational design of protein mu-

tants with enhanced solubility, J.Mol. Biol., 427, 478-

490, https://doi.org/10.1016/j.jmb.2014.09.026.

Publisher’s Note. Pleiades Publishing remains

neutral with regard to jurisdictional claims in published

maps and institutional affiliations. AI tools may have

been used in the translation or editing of this article.