Elena
Piletska
a,
Dana
Thompson
a,
Rebecca
Jones
a,
Alvaro Garcia
Cruz
a,
Marta
Poblocka
b,
Francesco
Canfarotta
a,
Rachel
Norman
c,
Salvador
Macip
bc,
Donald J. L.
Jones
d and
Sergey
Piletsky
*a
aChemistry Department, College of Science and Engineering, University of Leicester, Leicester, LE1 7RH, UK. E-mail: sp523@le.ac.uk
bMechanisms of Cancer and Aging Laboratory, Department of Molecular and Cell Biology, University of Leicester, Leicester, LE1 7RH, UK
cFoodLab, Faculty of Health Sciences, Universitat Oberta de Catalunya, 08018 Barcelona, Spain
dDepartment of Cancer Studies, RKCSB, University of Leicester, Leicester, LE2 7LX, UK
First published on 12th October 2022
Cellular senescence has proved to be a strong contributor to ageing and age-related diseases, such as cancer and atherosclerosis. Therefore, the protein content of senescent cells is highly relevant to drug discovery, diagnostics and therapeutic applications. However, current technologies for the analysis of proteins are based on a combination of separation techniques and mass spectrometry, which require handling large sample sizes and a large volume of data and are time-consuming. This limits their application in personalised medicine. An easy, quick and inexpensive procedure is needed for qualitative and quantitative analysis of proteins expressed by a cell or tissue. Here, we describe the use of the “snapshot imprinting” approach for the identification of proteins differentially expressed by senescent cells. Molecularly imprinted polymer nanoparticles (MIPs) were formed in the presence of whole cells. Following trypsinolysis, protein epitopes protected by complex with MIPs were eluted from the nanoparticles and analysed by LC-MS/MS. In this work, “snapshot imprinting” was performed parallel to a standard proteomic “shaving approach”, showing similar results. The analysis by “snapshot imprinting” identified three senescent-specific proteins: cell division cycle 7-related protein kinase, partitioning defective three homolog B and putative ATP-dependent RNA helicase DHX57, the abundance of which could potentially make them specific markers of senescence. Identifying biomarkers for the future elimination of senescent cells grants the potential for developing therapeutics for age-related diseases.
Acute senescence can be considered as beneficial for the body, as it aids tissue regeneration. It is a product of specific stress and has a specific target to assist in the process of wound or injury healing and development, such as embryogenesis.4,10,11 Also, senescence can act as a cell-autonomous tumour suppressor and can therefore prevent the proliferation of cancerous cells due to the release of tumour suppressor genes, e.g. p53 and p16INK4a. Chronic senescence, on the other hand, can be detrimental to health, leading to disease and tumour formation.10 Previous reports have revealed that senescence has a clear link with ageing, and with this, age-related diseases and pathologies such asdiabetes, kidney disease, atherosclerotic plaques in arteries, skin ulcers, arthritic joints, inflammation, sarcopenia, adiposity, neurogenesis, fibrosis and glaucoma.4,8,10
Targeted elimination of SNCs can relieve symptoms of ageing and prolong lifespan.9 Several strategies have already been developed in order to interfere with the detrimental effects of cellular senescence, such as selectively inducing the death of SNCs (by the use of drugs classed as “senolytics” or the immune-mediated clearance of SNCs), preventing senescence (with “senoblockers”12) and SASP neutralisation (with drugs called “senostatics” or “senomorphics”).13,14 Current senolytics are not sufficiently specific, but have been proven to somewhat increase the health-span and alleviate diseases in mice, such as atherosclerosis, osteoarthritis, cataracts, sarcopenia and cancer,15 and some may be of natural origin.16 Efforts to optimise selectivity have been made by designing second generation senolytics that consist of a senolytic drug and a targeting factor such as a nanoparticle or an antibody against a specific exposed epitope.17,18 This involves identifying proteins that are overexpressed on the surface of SNCs.11,19 This approach, however, requires understanding of the difference in membrane proteome (or “surfaceome”) of the normal and senescent cells and identification of protein targets that can be used for specific delivery of cytotoxic agents.
There are several ways to map the surface of SNCs for their epitopes. However, these are difficult and present drawbacks. An example of such a technique is X-ray diffraction analysis. Although it is the most precise method for epitope mapping, growing suitable crystals of membrane proteins is difficult and is marred by purity issues.16,17 Another example is the use of NMR, posing an advantage over X-ray diffraction by avoiding the need for crystals. However, this technique is of little use when it comes to mapping SNC surfaces, which can be explained by the fact that low metabolising SNCs will not possess a large enough excess of proteins on the cell surface that differentiate them from normal cells.18 For the same reason, generating antibodies against SNCs and using those for epitope mapping will be very difficult.
Cell surface mapping is a useful research technique in drug development and diagnostics.20,21 It consists of “shaving” a significant segment of a cell surface protein by digesting live, intact cells so the generated peptides can be analysed by LC-MS/MS. In the past, we have used the “shaving approach” for the identification of potential senescent biomarkers of the p16 pathway (manuscript in press). Although the results were successful, it is difficult to achieve a precise control of experimental conditions in the “shaving approach”, and for that reason, it cannot be readily applied in the clinical practice. Recently, we have described an experimental approach for using molecular imprinting to identify peptide sequences on the protein surface with potential antigenic properties.22 The method involves synthesis of MIP nanoparticles in the presence of whole protein, partial proteolysis of the protein bound to polymer, and subsequent sequencing of released peptides that were bound to the polymer. The important concept behind this principle relies on the assumption that MIPs synthesised in the presence of protein protect from proteolysis peptide sequences on the protein surface that are involved in MIP formation. This approach provides the possibility of identifying regions of the protein surface that have not yet been demonstrated to be antigenic in vivo, but which may offer improved affinity for natural and synthetic receptors such as antibodies, MIPs, aptamers, etc.
The same principle can be applied to take a “snapshot imprinting” of the cellular processes occurring in the cells, in order to map the topography of the cells and to identify peptide sequences and corresponding proteins that are expressed on the cell surface (Fig. 1).23 The obvious limitation to this procedure lies in the fact that cell populations might contain fractions representing different phases of the cell cycle. At the different phases, different proteins may be expressed to a higher degree. Hence, the ideal target for “snapshot imprinting” will be cells that are “frozen” in a perpetual phase of cell arrest, such as senescent cells. To prove this concept, MIPs were synthesised in the presence of cells undergoing p53-induced senescence. The proteins identified by snapshot imprinting for control and senescent cells were compared with the intention to identify abundant proteins characteristic to senescence that can be used as markers. The results of the “shaving approach” and “snapshot imprinting” were also compared, showing certain similarity of data obtained. We hope to use this knowledge in the future to identify and target senescent cells for diagnostic and therapeutic purposes, thus potentially delaying the onset of disease and aging.
Fig. 1 .Identification of membrane proteins in a surfaceome analysis by “snapshot imprinting”.23 |
The MIP synthesis, enzyme proteolysis and peptide sequencing were performed as described earlier.22 The monomeric mixture, comprised of NIPAm, TBAm, MBAA, AA and 3-aminopropyl methacrylamide, was optimised for protein imprinting.25,26 The synthesised nanoparticles are nontoxic to cultured cells, although the monomer mixture itself might induce stress.11,27,28 In our experiments, we did not notice cell lysis triggered by imprinting, which is essential for enriching fraction of membrane proteins in collected samples (data not shown). After approximately 1 hour of polymerisation, the non-polymerised monomers were washed away and cells/MIPs complexes treated with trypsin. After tryptic digestion, the MIP/epitope complexes were separated from unwanted cellular components and trypsin by centrifugal filtration and washing. Separated MIPs have average diameters of 170–180 nm (Table 1), which are significantly smaller than cells (5–20 μm). It is expected that each particle can be templated with only one epitope for a single protein. It is possible to describe cell imprinting as “freezing” exposed fragments of cell proteins in their complex with polymeric networks. This allows performing subsequent cell lysis without worrying whether cells remain intact during trypsinolysis. The epitope templates were removed from MIPs by heating. Peptide epitopes were eluted, concentrated and sequenced by LC/MS-MS. The epitopes sequences were analysed, providing information about types of the proteins exposed on cell membranes, and which contributed to the formation of MIPs, and their abundance. The LC-MS/MS results collected from three experiments with EJp53 cells are summarised in ESI Tables 1 and 2.†
Control cells | Z-Ave | PdI | Senescent cells | Z-Ave | PdI |
---|---|---|---|---|---|
Mean (nm) | 178.8 | 0.4 | Mean | 169.8 | 0.5 |
Std Dev (nm) | 6.0 | 0.1 | Std Dev | 11.2 | 0.1 |
The size of MIPs collected after elution of peptides was characterised using dynamic light scattering (DLS) and transmission electron microscopy (TEM) techniques. It was found that the hydrodynamic diameter of the MIPs in water was on average 128 nm ± 39 nm (polydispersity index—0.02). This size corresponds to the value reported for MIPs imprinted with proteins.29 The typical appearance of the MIPs under the TEM was spherical with their diameter approximately two times smaller than measured by DLS (Fig. 3).
The analysis of surface proteome by “snapshot imprinting” and by the “shaving approach” showed certain similarities. “Snapshot imprinting” allowed the identification of 112 proteins, versus 93 in the “shaving approach” (ESI Tables 1 and 2†). Approximately 33% of these proteins were identical for both approaches tested here. It is useful to notice that the concentration of peptide epitopes discovered in “snapshot imprinting” is substantially higher, by almost two orders of magnitude. This is probably a result of pre-concentration of peptides on the surface of MIP nanoparticles. The additional benefits of “snapshot imprinting” lies in its ability to link sequences of identified epitopes with their ability to generate MIP nanoparticles. This is in contrast to established protocols, e.g. the “shaving approach”, where information obtained about peptides/protein structure cannot be used for generating antibodies due to lack of correlation between abundance of proteins and their immunogenicity. These binders can be used for future elimination of senescent cells, granting the potential for the development of therapeutics for age-related diseases. In “snapshot imprinting” and also in the proteolytic “shaving approach” less than half of the identified proteins are actually membrane proteins. This phenomenon, observed in classical surfaceome research, is not completely understood, and could be due to the following possibilities: (i) cytoplasmic proteins come from cell lysis, thus contaminating the surfaceome fraction; (ii) cytoplasmic proteins have reached the surface by unspecified exporting/secretory machinery.30 Whether cytoplasmic proteins without any canonical secretion/exporting or retention signal are really translocated across the membrane is still unresolved, although there are many evidences both in prokaryotes and eukaryotes.31–34
As expected, uninduced (control) cells showed a large excess of identified proteins that were imprinted successfully. The proteins prevalent in the senescent cells are presented in Table 2. It is useful to notice that no reliable senescent markers were discovered by the “shaving approach”. The peptides belonging to two histone proteins, which are nuclear proteins but show a relative abundance in senescent cells, were measured with too large error margins (Table 2). This does not allow us to stipulate with confidence that these proteins are indeed present predominantly in senescent cells. The reliability of peptide measurement by “snapshot imprinting” is substantially higher. Three proteins were identified that show higher abundance in senescent cells by a factor of more than 5: cell division cycle 7-related protein kinase, partitioning defective 3 homolog B and putative ATP-dependent RNA helicase DHX57. The brief discussion of the role of these proteins in senescence is provided below.
Accession | Description | Senescent cells, fmol | Control cells, fmol |
---|---|---|---|
Snapshot imprinting | |||
B2R6V2; B7Z5H7; O00311; Q6JSD6 | Cell division cycle 7-related protein kinase | 5721 ± 234 | 0.00 |
A0A087WYL6; A0A087WZG6 | Partitioning defective 3 homolog B | 6748 ± 735 | 27 ± 16 |
A0A087X1S9; Q587I4; Q8TEW8 | |||
Q6P158; A0A087WZ11; B4DKW2 | Putative ATP-dependent RNA helicase DHX57 | 11675 ± 660 | 2150 ± 238 |
Shaving approach | |||
P04908; P0C0S8; P0C0S9; P20671; Q16777; Q3ZBX9; Q6FI13; Q7L7L0; Q93077; Q96KK5; Q99878; Q9BTM1; P0C0S4; P0C0S5; P16104; Q32LA7; Q71UI9; Q8IUE6; Q96QV6 | Histone H2A type 2-C | 17.2 ± 14 | 0.1 ± 0.1 |
O60814; P06899; P23527; P33778; P58876; P62807; P62808; Q16778; Q2M2T1; Q32L48; Q5QNW6; Q93079; Q99877; Q99879; P57053; Q96A08; Q99880; Q8N257 | Histone H2B type 1-B | 19 ± 15.2 | 1.7 ± 1.1 |
The cell division cycle 7-related protein kinase (CDC7) is a serine/threonine kinase that phosphorylates minichromosome maintenance protein 2 (MCM2) of the eukaryotic pre-replication complex.35 This protein is found in the nucleus of the cell and is implicated in cell division, cell-cycle checkpoint mechanisms, and cancer progression and its overexpression is present in different cancers.36 The activity of the catalytic subunit CDC7 is positively regulated by a complex formation with its activation subunit, the Dbf4 protein, also known as ASK (activator of S-phase kinase in human).37–39 The CDC7-ASK/Dbf4 complex also commonly referred to as the Dbf4 dependent kinase (or DDK) regulates the timing of DNA replication origin firing throughout S phase mainly by phosphorylation of MCM proteins, the major components of replicative helicase.40,41 The overexpression of CDC7 noted in our experiments could be directly caused by p53 expression,42 and this would have to be investigated using other cellular models of senescence. Of note, increased levels of CDC7 may not be per se a specific marker of senescence, since it is detected in different cancers.36
The partitioning defective 3 homolog B (PARD3B) is a membrane protein found to be much more abundant in SNCs. Atypical protein kinase C zeta (PKC zeta) forms a complex with PARD3 and PARD6 to regulate normal epithelial cell apico-basolateral polarization.43 The dissociation of the PKC zeta/PPARD3/PARD6 complex is essential for the disassembly of the tight/adherent junction and epithelial–mesenchymal transition (EMT) that is critical for tumour spreading.44 Suppression of PARD3 is associated with altered expression of genes regulating wound healing, cell apoptosis/death and cell motility, and particularly upregulation of MAP3K1 and fibronectin, which are known to contribute to cancer progression.43 Experimental evidence suggest that reduced expression of PKC zeta/Pard3/Pard6 contributes to EMT, invasion, and chemoresistance.45 It is possible that PARD3 protein can function as a senescence marker, although more experiments are needed to compare expression of PARD3 in senescent cells versus proliferating non-cancer cells, which also have increased level of PARD3 expression, as compared to cancer cells.46
The Putative ATP-dependent RNA helicase (DHX57) is part of the ATP-dependent RNA helicases, which are involved in diverse cellular functions such as RNA splicing, unwinding of a DNA or RNA helix, ribosome assembly, initiation of translation, spermatogenesis, embryogenesis, and cell growth and division. The precise mechanism and the substrates of these enzymes have not been defined. It is known that they play important roles in all types of processes in RNA metabolism and its expression might be related to biotic and abiotic stresses.47 DHX57 expression levels are also positively correlated with chronological age in the cerebellum.48 This supports our finding that this protein could be a marker of the presence of senescent cells.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d2na00424k |
This journal is © The Royal Society of Chemistry 2022 |