SARS-CoV-2 spike and Telomerase RNAs compared to arrive at an explanation for increased ageing in alveolar cells in severe COVID-19

In this letter we investigate if SARS-CoV-2 RNA is involved in the increased ageing of alveolar cells. Our in silico study is explorative. With the results we are able to outline experiments with AEC2 repair of bleomycin damaged alveolar cells. If AEC2 repair capability is diminished by spike RNA then perhaps this result provides a first step on a route to treat immortal lung cancer cells.


Introduction
In a recent study of COVID-19, a (statistical) relation was found between severity of COVID-19 illness and a decrease in length of peripheral blood lymphocyte telomeres 1 . COVID-19 is caused by the SARS-CoV-2 virus which introduces its single strand RNA via the ACE2 and TMPRSS2 receptors into the cell 2 . SARS-CoV-2 causes from mild flu-like symptoms in approximately 80% of the cases to a severe lung and multi-organic failure which can result in death of a significant percentage of patients (viz. Sanchez-Vazquez, 1).
Telomeres are chromosome ends to protect against rearrangement and the DNA broken strand repair system 3 of the cell. When cells divide and DNA is replicated, the telomeres become shorter 4 . This is a normal consequence of cellular division and is a molecular mechanism of cellular ageing 5 .
Here we will focus on alveolar epithelial cells (AEC) and note a certain parallelism between the effect of severe COVID-19 on the lungs and idiopathic pulmonary fibrosis (IPF).The first thing we may observe is that lung alveolar integrity is related to telomere length 6 and the activity of the enzyme telomerase. If the number of telomeres goes below the Hayflick limit, the cell enters the senescence and mortality stage 7 .
Telomerase is a ribonucleoprotein complex to maintain telomeres 8 . Telomerase is synthesized in stem / progenitor cells but its de novo synthesis in ordinary cells is suppressed. When an ordinary cell escapes mortality it becomes a cancer cell. Secondly, IPF is an illness with increased prevalence in advanced age with the hallmark of activation of AEC and epithelium driven accumulation of lung connective tissue 9 . The age factor suggests a role for telomeres and telomerase. Telomeres are therefore center stage here. Note e.g. also in early life, length of telomeres is dynamic 10 and e.g. telomere reduction in skin cells is caused by UV radiation 11 .
Thirdly there are two kinds of AEC. AEC1, responsible for oxygen processing and the progenitor cells AEC2. The latter produce surfactant and can transform, when necessary, into AEC1 12 . The ability to go from AEC2 to AEC1 is in need of telomerase (viz. Parra,10). We note that human AEC2 has ACE2 receptors so AEC2 is vulnerable to SARS-CoV-2 infection (viz. Sanchez-Vazquez, 1).
Telomerase is a reverse transcriptase. The RNA (hTR) of telomerase is an integral part of the enzyme and contains the template to telomeres it attaches processively at the 3' end of chromosomal DNA. The architecture of hTR in the complex 13 disables hijacking by alien RNA. Furthermore, there is a control mechanism for the incorporation of RNA in telomerase 14 . When a lot of similar to hTR alien RNA is present in the cell, the assembly of telomerase might be hampered. When such an AEC2 cell later turns into AEC1, it will have less telomere repair possibilities and have shorter telomeres.

Method
Here, ways to compare RNA sequences are designed to establish a distance measure for "similar". In this step the secondary structure of the hTR is of importance (viz. Zhang,13). We developed in our computational lab, a method based on multidimensional scaling descriptive statistics 15, 16 of similarities among objects. The basis of the scaling approach is a measure of similarity between object i and object j, in symbols, δi,j. The elementary objects here are the (NTs) nucleotides; A, U(T), C and G. In RNA the NTs are connected by ribose phosphor sugar repetitive elements. In the present analysis we will only look at the sequences containing the NTs to determine δi,j . Each NT at a certain position contribute to the δi,j.
The first characteristic is pairwise comparison of quantum Helium approximate wave function Ψ solutions of, Ψ = Ψ, and the Hamiltonian. I.e.
Each atom pair in the NT molecule is treated as though their outer electrons are in a Helium "atom" with Hamiltonian as given (viz. Geurdes 16). In the potential energy term, A further qualitative (dis)similarity is based on H bridges between the NTs. This concept is also employed to determine a matrix for second order configuration influence. The second order structure matrix is believed to hold the (pseudo)knots that are relevant to the architecture of the RNA in the telomerase complex. Obviously, the complementarity computations in the matrix overestimate the architectural form. However, the biochemical relevant architecture is present as a sub matrix. 12 1,2 1 2 In addition, per three NTs a four dimensional Euclidean distance computation was performed. Each NT represents a dimension in 4 space. Another point was a scaled qualitative categorization of similarity of three NTs in their amino acid effect. Finally, data from ATR-IR spectra (pubChem) were used for the 4 NTs to establish a similarity and to connect to a more semi-empiric set of data. Here axes of different NTs had to be transformed into each other in order to make the comparison. A normalization of ( 1 1+| − | ) was employed for the n-th and m-th NT in the two to be compared sequences.
Finally, RNA sequences are compared modulo the implicit restrictions and theoretical assumptions. The modeling details can be found in Geurdes,16. In the present case we first employed classical MDS and subsequently isoMDS in R. Then the second order configuration matrix was employed to the two axes and a subsequent isoMDS provides the result projection in a two dimensional space.
In this space the 75% Euclidean radius of the circle around the origin (0,0), i.e. , R75(00) , is a measure of similarity of the two RNA sequences. 75% of the points lie within the perimeter. Its rationale is that the coordinates sum to (0,0) in the projection. Furthermore, for computational convenience, the comparison in Fig. 1 is based upon 4 separate comparisons, each of size 271, with start points 1, 100, 199 and 268. Here, rotational freedom around the "out of plane" axis through (0,0) is employed to obtain the configuration with the smallest R75(00).

Material
In the in silico experiments we employed S spike data from GenBank: MT419837.1 and GenBank: U86046.1 for hTR.

Discussion 1
Before presenting the result of our computations, an alternative explanation is given. Fibrosis 2 can be caused by ACE2 blocking 18 . If the merging of the virus with the AEC2 membrane destroys the 3 ACE2 enzyme, it most likely will induce multiple divisions in order to maintain the ACE2 enzyme 4 function at a required level. Shortening of telomeres are expected in that case. Below we will explain 5 telomere shortening in severe COVID-19 as follows. 6 With the use of a 2 dimensional projection, the match with randomly generated RNA, Fig. 1  7 (B) & (D), for both S1 and S2 produces a larger R75(0,0) than with biological RNA. Further, S2 has the 8 lowest R75(0,0) value. Moreover, reversal of the direction of comparing 5'->3' S2 RNA with 3'->5' hTR 9 gave a relatively high R75(0,0), viz. Fig. 1 (E) vs (A). Modulo the assumptions in our model, we then b 10 can conclude that our in silico experiments indicate that the S2 part of the SARS-CoV-2 spike mRNA 11 (start at RNA position: 21563+2028=23591 is 1 S2 RNA) best fit the hTR of telomerase. Therefore, the 12 in vitro experiment that we propose is to have AEC2 cells that are able to assemble telomerase and 13 to introduce in those cells SARS-CoV-2 S2 spike RNA. 14 A possible experimental set up is in vitro bleomycin-induced lung epithelial cell (LEC)  15 apoptosis. Let us concentrate on AEC2 cells. Bleomycin causes an initial increase, and then a 16 reduction, in telomerase activity… 19 . The influence of S2 spike RNA on telomerase synthesis can 17 therefore be quantified by looking at the first 24 h telomerase increase in LEC of the ARC2 kind (viz. 18 Fridlender,19). We predict that S2 spike RNA treated AEC2 will produce less telomerase activity in 19 this peak of 24h. The reduction in the activity can be quantified with G-quadruplex-intercalating 20 porphyrin telomerase inhibitor. Further, the telomerase activity can also be quantified with PCR 21 based telomere repeat amplification protocol (TRAP) 20 . Detection of telomerase activity is briefly 22 presented in reference 19. An in vivo experiment can be modeled as an AEC2 transplantation 23 experiment 21 where AEC2 and S2 RNA treated AEC2 can give a difference in recovery grade from 24 bleomycine induced lung injury. 25 Finally we note that if the S2 RNA lowers the telomerase activity then a first step on the road 26 to a possible treatment of proliferation of immortal lung cells could be found. Telomerase activity 27 was found to be absent in most normal human somatic cells but present in over 90% of cancerous 28 cells (viz. Cong,7). Immortalization is the hallmark of malignant transformation and a premalignant 29 phenotype 22 . If S2 RNA has the predicted hampering effect on telomerase synthesis and can be 30 delivered to the cancer cells, then the immortality of this type of cells is destroyed. Perhaps that the 31 increase of caveolin-1 secretion of immortalized AEC2, can be a sign for a possible parasite vector 32 delivering the S2 RNA. 33