Pseudoknot (PK) is one of the most common structural motifs found in various functional RNAs, such as transfer RNA (tRNA), ribosomal RNA (rRNA), messenger RNA (mRNA), riboswitches and ribozymes[15–19]. PKs in these RNAs have crucial roles in the regulation of various biological processes such as translation, viral replication and catalytic activities, and thus, they can be a potential target in medicinal chemistry[15]. PK structures contain a minimum of two helical stems (S1 and S2), which are generally connected by single-stranded loops (L1 to L3) (Fig. 1)[16]. In the cases of H-type PKs, one of the simplest PK structures, their structural stabilities are closely linked with the folding-unfolding dynamics of one of the two helical regions that distinguish PK and hairpin (HP) conformers (Fig. 1). Previously, we analysed the stability of the S2 stem region in the H-type PK under physiological buffer conditions and demonstrated that the stability of the S2 stem region is predictable according to its sequence compositions based on the nearest-neighbour base pairs[20]. In the study, the absolute stabilities (−ΔGs) of the S2 stem regions were found to be lower than those of canonical RNA duplexes consisting of the same base pairs. The lower stabilities were considered due to conformational restriction and an alteration in the hydration state of the S2 stem region affected by the loop nucleotides connecting the S1 and S2 stem regions. However, to perform effectively their diverse biological functions, the PKs should be stable enough to a certain extent within the cells. Hence, the investigation of crowding effects on PK stability, including those caused by small chemicals and macromolecules, is crucial to understanding the mechanisms that are connected to their roles in biological systems. This process is also essential for predicting the PK-forming regions in the RNA transcripts that can be targeted to construct therapeutic materials.
Table 1. Sequences of PK RNA oligonucleotides used in this work.
RNA Name
|
RNA Sequences (5´ to 3´) a
|
PK-1
|
GGCGCAGUGUCGCAGCGCCACUCAAAAGCGACA
|
PK-2
|
GGCGCAGUGCGUCAGCGCCACUCAAAAGACGCA
|
PK-L1AB2-1
|
GGCGCXXUGUCGCAGCGCCACUCAAAAGCGACA
|
PK-L1AB3-1
|
GGCGCXXXUGUCGCAGCGCCACUCAAAAGCGACA
|
a. Red, blue and green colours denote the S1 stems, and S2 stems and incorporated abasic site instead of L1 nucleobases, respectively.
In this study, we used a well-known molecular crowding agent, polyethylene glycol (PEG), with different molecular weights of 200 (PEG200) and 8000 (PEG8000), as a model of small chemicals and macromolecules, respectively, for inducing the crowding environments[21, 22]. Since the formation of the S2 stem region is essential to distinguish the PK structure from a conventional HP structure, the structural stability of PK can be assessed by evaluating the thermal denaturation of the S2 stem region (Figure S1)[20]. We have selected two PKs, PK-1 and PK-2 (Table 1), to be analysed under the crowding environments. These PKs have modified sequences of natural PK derived from mouse mammary tumour virus (MMTV), in which S2 stem regions have the same nearest neighbour base pairs and thus show similar stability[20]. PK-forming RNA oligonucleotides, which have abasic deoxynucleotides in the L1 region (PK-L1AB2-1 and PK-L1AB3-1), were designed to evaluate the effects of the loop nucleobases on PK stability (Table 1). RNA oligonucleotides, which only form HP conformer having the same base pairs with the S1 stem region of the corresponding PK (Table S1), were also designed to extract the melting behaviours of the S2 stem region following our previous strategy[20].
Circular dichroism (CD) spectra of PK-1, PK-2, and HP-1 were analysed in a buffer containing 10 mM Na2HPO4 (pH 7), 100 mM NaCl, and 1 mM Na2EDTA in the absence and presence of 30 wt% PEG200 or 20 wt% PEG8000 at 10°C (Figure S2). The spectra of PKs and HP in both dilute and crowded buffer conditions showed almost the same profiles, with two negative peaks around 210 and 237 nm and one positive peak around 265 nm. These peaks are characteristics of the CD spectra for a typical RNA duplex[23]. Therefore, the results suggest that the PK and HP, which both contain the stem duplex as a main structural motif, were similarly formed regardless of the additional cosolutes. As the CD signal at 210 nm showed the maximum difference between PK and HP (Figure S3), the thermal denaturation profiles of the S2 stem region of PKs were evaluated by recording ellipticity signals at 210 nm after subtracting the signal obtained by HP (Figure S1).
Thermal denaturation profiles of the S2 stem region of PK-1 and PK-2 were evaluated in the absence and presence of different concentrations of PEG200 (0 to 30%) and PEG8000 (0 to 20%). Typical denaturation profiles of PK-1, PK-2, and HP-1 were shown in Figure S4, and normalized melting profiles of PK-1 and PK-2 after subtraction of the signal obtained from HP-1 were shown in Fig. 2 and Figure S5, respectively. Thermodynamic parameters, ΔH°, TΔS°, and ΔG° at 37 °C (ΔG°37), for the formation of the S2 stem regions, were calculated from the melting profiles by fitting an equation used for the two-state conformational transition (Table 2).
A decrease of water activity in crowding environment induced by small cosolutes such as PEG200 destabilizes canonical duplexes in both DNA and RNA[10, 12]. Accordingly, in Figure S4, ellipticity increases over 70°C, which corresponds to the dissociation of the S1 stem region in PKs or HP-1, showed a lower melting temperature in the presence of 30 wt% PEG200 compared to the absence of cosolute. In contrast to the destabilization of the S1 stem, stability in the S2 stem of both PK-1 and PK-2 was increased with the increasing PEG200 concentration (Table 2). The differences in ΔG°37 values (ΔΔG°37) in the absence and presence of 30 wt% PEG200 were − 0.9 and − 1.1 kcal mol− 1 for PK-1 and PK-2, respectively. The opposite effects on the stability of the S2 and S1 regions provided by PEG200 suggested that the S2 stem region has a hydration state different from the S1 stem region, which is expected to be less hydrated. PEG200 is also known to decrease the dielectric constant of the solution[10, 24]. The decreased dielectric constant in the presence of PEG200 also has the potential to stabilize the S2 stem region by enhancing cation (here, sodium salt) binding to the negatively charged phosphate backbone in the PK structure that results in less intra-strand repulsion.
Table 2. Thermodynamic parameters for the formation of S2 in different PK structures in the presence and absence of PEG200 and PEG8000 which were calculated by CD melting studies collected at 210 nm in a buffer containing 10 mM Na2HPO4 (pH 7), 100 mM NaCl, and 1 mM Na2EDTA.
PK Name
|
Molecular weight
of PEG
|
PEG
(wt%)
|
ΔH° a
(kcal mol-1)
|
TΔS° a
(kcal mol-1)
|
ΔG°37 a
(kcal mol-1)
|
Tm a
(°C)
|
ΔΔG°37
ΔG°37 (with PEG) − ΔG°37 (without PEG)
(kcal mol-1)
|
PK-1
|
200
|
0
|
-38.3 ± 1.7
|
-37.6 ± 1.8
|
-0.7 ± 0.0
|
42.4 ± 0.5
|
–
|
10
|
-31.5 ± 1.8
|
-30.6 ± 1.8
|
-0.9 ± 0.0
|
46.0 ± 0.8
|
-0.2
|
20
|
-35.6 ± 2.0
|
-34.5 ± 1.9
|
-1.1 ± 0.1
|
47.0 ± 0.1
|
-0.4
|
30
|
-39.5 ± 4.6
|
-37.9 ± 4.4
|
-1.6 ± 0.2
|
50.0 ± 0.7
|
-0.9
|
PK-2
|
200
|
0
|
-32.6 ± 1.1
|
-32.0 ± 1.2
|
-0.6 ± 0.0
|
43.0 ± 0.4
|
–
|
10
|
-30.2 ± 1.3
|
-29.3 ± 1.2
|
-0.9 ± 0.1
|
46.8 ± 0.2
|
-0.3
|
20
|
-35.8 ± 2.6
|
-34.7 ± 2.5
|
-1.2 ± 0.0
|
47.4 ± 0.6
|
-0.6
|
30
|
-42.1 ± 5.4
|
-40.4 ± 5.2
|
-1.7 ± 0.2
|
50.1 ± 0.6
|
-1.1
|
PK-L1AB2-1
|
200
|
0
|
-37.2 ± 6.4
|
-35.5 ± 6.3
|
-1.7 ± 0.2
|
52.1 ± 1.5
|
–
|
10
|
-34.5 ± 3.4
|
-32.7 ± 3.3
|
-1.8 ± 0.2
|
54.2 ± 0.9
|
-0.1
|
20
|
-39.4 ± 2.1
|
-37.3 ± 2.0
|
-2.1 ± 0.1
|
54.1 ± 0.6
|
-0.4
|
30
|
-40.7 ± 1.1
|
-38.7 ± 1.1
|
-2.0 ± 0.0
|
53.4 ± 0.1
|
-0.3
|
PK-L1AB3-1
|
200
|
0
|
-36.1 ± 4.8
|
-34.6 ± 4.6
|
-1.4 ± 0.1b
|
49.8 ± 0.4
|
–
|
10
|
-39.1 ± 4.1
|
-37.3 ± 3.9
|
-1.8 ± 0.2
|
51.9 ± 0.5
|
-0.4
|
20
|
-38.9 ± 4.5
|
-37.1 ± 4.3
|
-1.8 ± 0.2
|
51.8 ± 0.5
|
-0.4
|
30
|
-38.7 ± 1.6
|
-36.8 ± 1.5
|
-1.9 ± 0.1
|
53.3 ± 0.2
|
-0.5
|
PK-1
|
8000
|
0
|
-38.3 ± 1.7
|
-37.6 ± 1.8
|
-0.7 ± 0.0
|
42.4 ± 0.5
|
–
|
5
|
-31.4 ± 2.8
|
-30.6 ± 2.7
|
-0.8 ± 0.2
|
45.2 ± 1.4
|
-0.1
|
10
|
-41.2 ± 1.5
|
-39.9 ± 1.6
|
-1.4 ± 0.1b
|
47.6 ± 1.3
|
-0.7
|
15
|
-37.7 ± 2.6
|
-36.2 ± 2.5
|
-1.5 ± 0.1
|
49.6 ± 0.7
|
-0.8
|
20
|
-40.2 ± 8.8
|
-38.2 ± 8.5
|
-1.9 ± 0.3b
|
52.7 ± 1.3
|
-1.2
|
PK-2
|
8000
|
0
|
-32.6 ± 1.1
|
-32.0 ± 1.2
|
-0.6 ± 0.0
|
43.0 ± 0.4
|
–
|
5
|
-30.1 ± 2.8
|
-29.2 ± 2.8
|
-0.9 ± 0.0
|
46.7 ± 0.8
|
-0.3
|
10
|
-37.3 ± 4.1
|
-36.0 ± 3.9
|
-1.3 ± 0.2
|
48.4 ± 0.7
|
-0.7
|
15
|
-31.9 ± 1.3
|
-30.5 ± 1.2
|
-1.5 ± 0.0
|
51.9 ± 0.4
|
-0.9
|
20
|
-49.4 ± 4.6
|
-47.1 ± 4.4
|
-2.4 ± 0.2
|
52.6 ± 0.3
|
-1.8
|
PK-L1AB2-1
|
8000
|
0
|
-37.2 ± 6.4
|
-35.5 ± 6.3
|
-1.7 ± 0.2
|
52.1 ± 1.5
|
–
|
5
|
-45.9 ± 3.3
|
-43.7 ± 3.3
|
-2.2 ± 0.1
|
52.9 ± 0.9
|
-0.5
|
10
|
-51.0 ± 7.4
|
-48.2 ± 7.0
|
-2.7 ± 0.4b
|
54.6 ± 0.6
|
-1.0
|
15
|
-51.7 ± 6.4
|
-48.6 ± 6.0
|
-3.1 ± 0.4
|
56.9 ± 0.9
|
-1.4
|
20
|
-51.1 ± 3.1
|
-47.7 ± 2.8
|
-3.4 ± 0.3
|
58.9 ± 0.9
|
-1.7
|
PK-L1AB3-1
|
8000
|
0
|
-36.1 ± 4.8
|
-34.6 ± 4.6
|
-1.4 ± 0.1b
|
49.8 ± 0.4
|
–
|
5
|
-48.9 ± 3.7
|
-46.5 ± 3.6
|
-2.4 ± 0.1
|
52.8 ± 0.6
|
-1.0
|
10
|
-51.9 ± 3.3
|
-49.1 ± 3.2
|
-2.7 ± 0.2b
|
54.1 ± 0.9
|
-1.3
|
15
|
-49.8 ± 2.3
|
-46.9 ± 2.2
|
-2.9 ± 0.1
|
55.9 ± 0.6
|
-1.5
|
20
|
-49.2 ± 4.2
|
-46.2 ± 3.9
|
-3.0 ± 0.4
|
57.1 ± 0.9
|
-1.6
|
a Values and errors are the average ± SD of at least three or more experiments.
b Slight deviation in averaged ΔG°37 values compared to the difference of ΔH° and TΔS° values in the left columns is due to the rounding off process of each replicated data.
To gain insight into the stabilization of the S2 stem in the presence of PEG200, the stability of mutated PK, which contains abasic sites in the L1 region of PK-1, PK-L1AB2-1 and PK-L1AB3-1, was evaluated in the presence of PEG200 (Table 2). The S2 stem region of PK-L1AB2-1 and PK-L1AB3-1 was more stable than that of PK-1 and PK-2 both in the absence and presence of PEG200, suggesting the nucleobases at the L1 region destabilized the S2 stem of the PKs. However, the degree of the destabilization provided by the L1 nucleobases became smaller in the presence of a higher concentration of PEG200, as the negative slope of ΔG°37 values depending on the PEG200 concentration was higher for PK-1 and PK-2 compared to PK-L1AB2-1 and PK-L1AB3-1 (Figure S6A). The results suggest that the nucleobases in the L1 region disturbed the hydration network, which is an important factor for stabilizing the duplex, around the S2 stem region in PK-1 and PK-2 and destabilized the S2 stem in the condition of diluted buffer solution. The reduced number of water molecules around the S2 stem region was in turn responsible for the observed stabilization under the crowding environment, where the water activity is less than the diluted buffer solution. Additionally, the phosphate and ribose backbones at the L1 region also have the potential to disturb the hydration network, which is manifested by the slight stabilization of the S2 stem region of PK-L1AB2-1 and PK-L1AB3-1 in the presence of PEG200, where ΔΔG°37 of PK-L1AB2-1 and PK-L1AB3-1 in the absence and presence of 30 wt% PEG200 was − 0.3 and − 0.5 kcal mol− 1, respectively (Table 2). Another plausible mechanism for this slight stabilization could be due to the enhanced interaction between phosphate and cation under the decreased dielectric constant in the crowding environment, which reduced the electrostatic repulsion.
In the case of a crowding environment induced by PEG8000, the S1 stem in PKs and HP-1 was not affected much compared to the destabilization caused by PEG200 (Figure S4), as it is expected from the insensitivity of the duplex stability to PEG8000[25]. Compared to the S1 region, the S2 region of PKs was stabilized in the presence of PEG8000 (Figs. 2 and S5). ΔΔG°37 in the absence and presence of 20 wt% PEG8000 were − 1.2 and − 1.8 kcal mol− 1 for PK-1 and PK-2, respectively (Table 2). The stabilization provided by PEG8000 was higher than PEG200 at the same wt% concentration. The stabilization of RNA tertiary structures has often been observed in the presence of large cosolute molecules attributed to the excluded volume because the folded state of the RNAs tends to be smaller than their denatured form. It is considered that the folded PK is more compact than its hairpin conformer with a dissociated S2 stem region and is a favourable conformer in the presence of PEG8000. PK-L1AB2-1 and PK-L1AB3-1 also exhibited significant S2 stem stabilization in the presence of PEG8000 to a similar extent as PK-1 and PK-2 (Figure S6B). The results suggest that the nucleobases in the L1 region do not contribute to the stabilization of PK attributed to the excluded volume, because nucleobases are relatively small compared to the overall PK structure.
In summary, we have explored the structural stability of PK structures under crowding environments with small and large cosolutes. Although both cosolutes stabilized the S2 stem region, the stabilization mechanisms were different. The smaller cosolute (PEG200) stabilized the S2 stem region by reducing the water activity in the solution. The reduced hydration of the stem region was expected due to the presence of nucleobases in the L1 loop region, i.e., the stabilization effect originated in the local conformational state. On the other hand, a larger cosolute (PEG8000) stabilized the S2 stem region by providing excluded volume. Since the size of the molecule is important to consider the excluded volume effect the stabilization provided by the PEG8000 originated in the global PK structure rather than the microenvironment of the local regions (Fig. 3).
The extent of molecular crowding provided by both small and large molecules changes depending on not only the different regions of cells such as the nucleus, nucleolus, and cytosol, but also the timing in cell activity such as replication cycle, differentiation, and ageing[27, 28]. In addition, the biological functions of the PKs independently depend on not only their global stability but also local conformational states by considering further interactions between PKs and other biomolecules, such as proteins and metabolites. The differences in the stabilizing effects of molecular crowding clarified in this study provide critical insight and serve as a bridge between the diverse structural characteristics of PKs and mechanisms of their functionalization in varying cellular contexts. Moreover, recognition of the nuanced impact of different cosolute molecules on PK stability and function also holds promise for future medicinal chemistry aimed at exploiting PKs as therapeutic targets.
This work was supported by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) and Japan Society for the Promotion of Science (JSPS) (KAKENHI Grant No. 21H05108 and 23H02087), especially for Grant-in-Aid for Scientific Research (S) (22H04975), JSPS Core-to-Core Program (JPJSCCA20220005), The Hirao Taro Foundation of KONAN GAKUEN for Academic Research, and the Chubei Itoh Foundation. We thank Eri Miyaura for her help with experiments.